It is now well established that a large fraction of stars forms along molecular filaments. However, the origin of the observed filament properties, as well as their formation, evolution, and fragmentation, are not well understood yet. Indeed, conventional theoretical models predict that thermally supercritical filaments (line mass > 2cs²/G ≈ 17 M☉ pc⁻¹) undergo radial collapse due to insufficient thermal support, suppressing fragmentation. Observations, however, show that even high line mass filaments with Mline ≫ Mline,crit form stars. Previous theoretical studies attempted to solve this discrepancy with HD and ideal-MHD simulations of accreting filaments, with both homogenous and non-homogenous turbulent flows, but the turbulence driven inside the filament was insufficient to support the filament against radial collapse. In our new study (Abe et al. 2025, ApJ), we use non-ideal MHD simulations including ambipolar diffusion of filament formation and evolution through accretion and identify a new mechanism that generates non-thermal pressure within the filament that counteracts gravity preventing the radial collapse of a massive filament. We name this mechanism “STORM » (Slow-shock-mediated Turbulent flOw Reinforced by Magnetic diffusion) where the corrugations developing at the surface of the filament by the slow-shock instability create over-densities, or blobs, which detach and inject turbulence within the filament. Hence, the turbulence within the filaments is enhanced, rather than suppressed, by ambipolar diffusion—an effect counterintuitive to the usual role of ambipolar diffusion in smoothing structures. We find that the interaction between magnetic diffusion and the slow-shock instability efficiently injects turbulent ram pressure that maintains the filament width against collapse. Our results show that even massive filaments with a line mass ≈ 100 M☉ pc⁻¹ can retain a width of ≈ 0.1 pc through this mechanism, in agreement with observations. New observational results with ALMA (Tachihara et al. 2024, ApJ) and JWST (Andre et al. 2025, ApJL) seem to show evidence of corrugations and blob-like structures supporting our simulations and theoretical interpretation. Our findings not only offer a promising theoretical understanding to the observed width of massive star forming filaments but may also provide new insights into the properties of cores, such as their mass and angular momentum.
Contact local : Thierry FOGLIZZO
Organisation : Frédéric GALLIANO
DAp