We extend our previous work on simulations with the code RAMSES on accretion driven turbulence by including self-gravity and study the effects of core formation and collapse. We show that radial accretion onto filaments drives turbulent motions which are not isotropic but radially dominated. In contrast to filaments without gravity, the velocity dispersion of self-gravitating filaments does not settle in an equilibrium. Despite showing similar amounts of driven turbulence, they continually dissipate their velocity dispersion until the onset of core formation. This difference is connected to the evolution of the radius as it determines the dissipation rate. In the non-gravitational case filament growth is not limited and its radius grows linearly with time. In contrast, there is a maximum extent in the self-gravitational case resulting in an increased dissipation rate. Furthermore, accretion driven turbulence shows a radial profile which is anti-correlated with density. This leads to a constant turbulent pressure throughout the filament. As the additional turbulent pressure does not have a radial gradient it does not contribute to the stability of filaments and does not increase the critical line-mass. However, this radial turbulence does affect the radius of a filament, adding to the extent and setting its maximum value. Moreover, the radius evolution also affects the growth timescale of cores which compared to the timescale of collapse of an accreting filament limits core formation to high line-masses.

We study accretion driven turbulence for different inflow velocities in star forming filaments using the code ramses. Filaments are rarely isolated objects and their gravitational potential will lead to radially dominated accretion. In the non-gravitational case, accretion by itself can already provoke non-isotropic, radially dominated turbulent motions responsible for the complex structure and non-thermal line widths observed in filaments. We find that there is a direct linear relation between the absolute value of the total density weighted velocity dispersion and the infall velocity. The turbulent velocity dispersion in the filaments is independent of sound speed or any net flow along the filament. We show that the density weighted velocity dispersion acts as an additional pressure term supporting the filament in hydrostatic equilibrium. Comparing to observations, we find that the projected non-thermal line width variation is generally subsonic independent of inflow velocity.

Observations of prestellar cores in star-forming filaments show two distinct morphologies. While molecular line measurements often show broad cores, submillimeter continuum observations predominantly display pinched cores compared to the bulk of the filament gas. In order to explain how different morphologies arise, we use the gravitational instability model where prestellar cores form by growing density perturbations. The radial extent at each position is set by the local line-mass. We show that the ratio of core radius to filament radius is determined by the initial line-mass of the filament. Additionally, the core morphology is independent of perturbation length scale and inclination, which makes it an ideal diagnostic for observations. Filaments with a line-mass of less than half its critical value should form broad cores, whereas filaments with more than half its critical line-mass value should form pinched cores. For filaments embedded in a constant background pressure, the dominant perturbation growth times significantly differ for low and high line-mass filaments. Therefore, we predict that only one population of cores is present if all filaments within a region begin with similar initial perturbations.

We present a solution for the observed core fragmentation of filaments in the Taurus L1517 dark cloud which previously could not be explained (Hacar et. al 2011). Core fragmentation is a vital step for the formation of stars. Observations suggest a connection to the filamentary structure of the cloud gas but it remains unclear which process is responsible. We show that the gravitational instability process of an isothermal cylinder can account for the exhibited fragmentation under the assumption that the perturbation grows on the dominant wavelength. We use numerical simulations with the code RAMSES, estimate observed column densities and line-of-sight velocities and compare them to the observations. A critical factor for the observed fragmentation is that cores grow by redistributing mass within the filament and thus the density between the cores decreases over the fragmentation process. This often leads to wrong dominant wavelength estimates as it is strongly dependent on the initial central density. We argue that non-linear effects also play an important role on the evolution of the fragmentation. Once the density perturbation grows above the critical line-mass, non-linearity leads to an enhancement of the central core density in comparison to the analytical prediction. Choosing the correct initial conditions with perturbation strengths of around 20\%, leads to inclination corrected line-of-sight velocities and central core densities within the observational measurement error in a realistic evolution time.

We study the stability of filaments in equilibrium between gravity and internal as well as external pressure using the grid based AMR-code RAMSES. A homogeneous, straight cylinder below a critical line mass is marginally stable. However, if the cylinder is bent, e.g. with a slight sinusoidal perturbation, an otherwise stable configuration starts to oscillate, is triggered into fragmentation and collapses. This previously unstudied behavior allows a filament to fragment at any given scale, as long as it has slight bends. We call this process `geometrical fragmentation'. In our realization the spacing between the cores matches the wavelength of the sinusoidal perturbation, whereas up to now, filaments were thought to be only fragmenting on the characteristical scale set by the mass-to-line ratio. Using first principles, we derive the oscillation period as well as the collapse timescale analytically. To enable a direct comparison with observations, we study the line-of-sight velocity for different inclinations. We show that the overall oscillation pattern can hide the infall signature of cores.

Short version: We apply a new synergetic radiative transfer method: the derived 1D density profiles are both consistent with a cut through the Herschel PACS/SPIRE and JCMT SCUBA-2 continuum maps of L1689B and with a derived local interstellar radiation field. Choosing an appropriate cut along the filament major axis, we minimize the impact of the filament emission on the modeling. For the bulk of the core (5000-20000 au) an isothermal sphere model with a temperature of around 10 K provides the best fits. We show that the power law index of the density profile, as well as the constant temperature can be derived directly from the radial surface brightness profiles. For the inner region (< 5000 au), we find a range of densities and temperatures that are consistent with the surface brightness profiles and the local interstellar radiation field. Based on our core models, we find that pixel-by-pixel single temperature spectral energy distribution fits are incapable of determining dense core properties. We conclude that, to derive physical core properties, it is important to avoid an azimuthal average of core and filament. Correspondingly, derived core masses are too high since they include some mass of the filament, and might introduce errors when determining core mass functions. The forward radiative transfer methods also avoids the loss of information owing to smearing of all maps to the coarsest spatial resolution. We find the central core region to be colder and denser than estimated in recent inverse radiative transfer modeling, possibly indicating the start of star formation in L1689B.