SHOCK PROPAGATION AND THE GENERATION OF MAGNETOHYDRODYNAMIC WAVE-FIELDS IN INHOMOGENEOUS MOLECULAR CLOUDS

We develop a simple one-dimensional model for the interaction of a ste ady, thin, planar shock wave with a nonrigid cloud which may be in mot ion relative to the surrounding medium, and we apply the model to shoc ks impinging on, and propagating through, molecular clouds. Both ''adi abatic'' (gamma = 5/3) and radiative (gamma = 1) shocks are considered and we allow for the presence of a uniform magnetic field directed ei ther parallel or perpendicular to the shock normal. The former field o rientation is equivalent to the hydrodynamic case, and the latter invo lves only fast MHD shocks. We focus on the manner in which such shocks can generate internal kinetic motions in the cloud on a range of size and density scales through the direct acceleration of cores and clump s by shocks transmitted into them and through the generation of an MHD wavefield via the reflection of the incident shock at clump boundarie s. We find that stronger incident Mach numbers and smaller density con trasts lead to more efficient cloud acceleration, as do isothermal int ercloud shocks and small intercloud magnetic held strengths. The accel eration efficiency is insensitive to the adiabatic index and the magne tic field strength in the cloud itself. For typical parameter choices, the direct acceleration of clouds and clumps by strong shocks is foun d to be substantial and could at least in part account for their obser ved velocity dispersions. If the shocks are moderately weak, the final velocity of the cloud is linearly related to its initial velocity, wi th higher acceleration giving shallower slopes (i.e., final velocity d istributions which are less sensitive to the initial distribution). Co mpared to the kinetic energy of the postshock cloud, the energy given to the wavefield at each encounter is small, and the heating of the in terclump medium by the dissipation of this wavefield is found to be in sufficient to balance the cooling rate in the cloud as a whole (althou gh it may be important in particular regions), even if this medium is warm, unless it is also extremely tenuous (n less than or similar to 0 .1 cm(-3)). Nevertheless, the correction for the velocity imparted to the cloud leads to a substantial increase in the critical incident Mac h number for wave emission over that reported by Spitzer for the rigid case. The implications of our model for shock-induced star formation are discussed briefly.