We present the results of smoothed particle hydrodynamic simulations investigating the evolution and fragmentation of filaments that are accreting from a turbulent medium. We show that the presence of turbulence, and the resulting inhomogeneities in the accretion flow, play a significant role in the fragmentation process. Filaments which experience a weakly turbulent accretion flow fragment in a two-tier hierarchical fashion, similar to the fragmentation pattern seen in the Orion Integral Shaped Filament. Increasing the energy in the turbulent velocity field results in more sub-structure within the filaments, and one sees a shift from gravity-dominated fragmentation to turbulence-dominated fragmentation. The sub-structure formed in the filaments is elongated and roughly parallel to the longitudinal axis of the filament, similar to the fibres seen in observations of Taurus, and suggests that the fray and fragment scenario is a possible mechanism for the production of fibres. We show that the formation of these fibre-like structures is linked to the vorticity of the velocity field inside the filament and the filament's accretion from an inhomogeneous medium. Moreover, we find that accretion is able to drive and sustain roughly sonic levels of turbulence inside the filaments, but is not able to prevent radial collapse once the filaments become supercritical. However, the supercritical filaments which contain fibre-like structures do not collapse radially, suggesting that fibrous filaments may not necessarily become radially unstable once they reach the critical line-density.

Context. We present an initial overview of the filamentary structure in the Orion A molecular cloud utilizing a high angular and velocity resolution C18O(1–0) emission map that was recently produced as part of the CARMA-NRO Orion Survey. Aims. The main goal of this study is to build a credible method to study varying widths of filaments which has previously been linked to star formation in molecular clouds. Due to the diverse star forming activities taking place throughout its ~20 pc length, together with its proximity of 388 pc, the Orion A molecular cloud provides an excellent laboratory for such an experiment to be carried out with high resolution and high sensitivity. Methods. Using the widely-known structure identification algorithm, DisPerSE, on a three-dimensional (PPV) C18O cube, we identify 625 relatively short (the longest being 1.74 pc) filaments over the entire cloud. We studied the distribution of filament widths using FilChaP, a python package that we have developed and made publicly available. Results. We find that the filaments identified in a two square-degree PPV cube do not overlap spatially, except for the complex OMC-4 region that shows distinct velocity components along the line of sight. The filament widths vary between 0.02 and 0.3 pc depending on the amount of substructure that a filament possesses. The more substructure a filament has, the larger is its width. We also find that despite this variation, the filament width shows no anticorrelation with the central column density which is in agreement with previous Herschel observations.

We use Smoothed Particle Hydrodynamic simulations of cold, uniform density, self-gravitating filaments, to investigate their longitudinal collapse timescales; these timescales are important because they determine the time available for a filament to fragment into cores. A filament is initially characterised by its line-mass, µ O , its radius,, and its aspect ratio,The gas is only allowed to contract longitudinally, i.e. parallel to the symmetry axis of the filament (the z-axis). Pon et al. (2012) have considered the global dynamics of such filaments analytically. They conclude that short filaments (A O 5) collapse along the z-axis more-or-less homologously, on a timescale t HOM ∼ 0.44 A O (Gρ O ) −1/2 ; in contrast, longer filaments (A O 5) undergo enddominated collapse, i.e. two dense clumps form at the ends of the filament and converge on the centre sweeping up mass as they go, on a time-scale t END ∼ 0.98Our simulations do not corroborate these predictions. First, for all A O 2, the collapse time satisfies a single equationwhich for large A O is much longer than the Pon et al. prediction. Second, for all A O 2, the collapse is end-dominated. Third, before being swept up, the gas immediately ahead of an end-clump is actually accelerated outwards by the gravitational attraction of the approaching clump, resulting in a significant ram pressure. For high aspect ratio filaments the end-clumps approach an asymptotic inward speed, due to the fact that they are doing work both accelerating and compressing the gas they sweep up. Pon et al. appear to have neglected the outward acceleration and its consequences.

We use smoothed particle hydrodynamic simulations to investigate the growth of perturbations in infinitely long filaments as they form and grow by accretion. The growth of these perturbations leads to filament fragmentation and the formation of cores. Most previous work on this subject has been confined to the growth and fragmentation of equilibrium filaments and has found that there exists a preferential fragmentation length scale which is roughly 4 times the filament's diameter. Our results show a more complicated dispersion relation with a series of peaks linking perturbation wavelength and growth rate. These are due to gravo-acoustic oscillations along the longitudinal axis during the sub-critical phase of growth. The positions of the peaks in growth rate have a strong dependence on both the mass accretion rate onto the filament and the temperature of the gas. When seeded with a multi-wavelength density power spectrum there exists a clear preferred core separation equal to the largest peak in the dispersion relation. Our results allow one to estimate a minimum age for a filament which is breaking up into regularly spaced fragments, as well as an average accretion rate. We apply the model to observations of filaments in Taurus by Tafalla & Hacar (2015) and find accretion rates consistent with those estimated by Palmeirim et al. (2013).

Molecular-line observations of filaments in star-forming regions have revealed the existence of elongated coherent features within the filaments; these features are termed fibres. Here we caution that, since fibres are traced in PPV space, there is no guarantee that they represent coherent features in PPP space. We illustrate this contention using simulations of the growth of a filament from a turbulent medium. Synthetic C 18 O observations of the simulated filaments reveal the existence of fibres very similar to the observed ones, i.e. elongated coherent features in the resulting PPV data-cubes. Analysis of the PPP data-cubes (i.e. 3D density fields) also reveals elongated coherent features, which we term sub-filaments. Unfortunately there is very poor correspondence between the fibres and the sub-filaments in the simulations. Both fibres and sub-filaments derive from inhomogeneities in the turbulent accretion flow onto the main filament. As a consequence, fibres are often affected by line-of-sight confusion. Similarly, sub-filaments are often affected by large velocity gradients, and even velocity discontinuities. These results suggest that extreme care should be taken when using velocity coherent features to constrain the underlying substructure within a filament.