We discuss partitioned time-integration of dynamic multiphysics problems, which commonly exhibit a multiscale behavior, requiring independent time-grids. The ideal method for solving these problems allows independent and adaptive time-grids, higher order time-discretizations, is fast and robust, and allows the coupling of existing subsolvers, executed in parallel. Waveform relaxation (WR) methods can potentially have all of these properties. These iterate on continuous-in-time interface functions, obtained via suitable interpolation. This way, one can allow for time adaptivity in the subsolvers. An important issue is to find suitable convergence acceleration, which is required for the iteration to converge quickly. We present a fast and highly robust, second order in time, adaptive WR method. Basis is a Dirichlet-Neumann coupling at the interface. The method uses an analytically optimal relaxation parameter derived for the fully-discrete scheme in 1D. This parameter depends on discretization and problem, and in practice leads to very fast linear convergence. A more black box alternative is the Quasi-Newton Waveform relaxation method presented for equidistant but different time grids. We discuss an extension of this method to the time adaptive case, with particular attention to the choice of the tolerances for the Waveform iteration, time adaptivity, and the inner iterative solvers. Numerical results demonstrate the robustness of both approaches.