Publications

We developed a computationally-efficient algorithm for performing the Helmholtz-Hodge decomposition (HHD) on velocity fields defined on block-based AMR grids, addressing the complexity posed by non-uniform grids. The method reformulates the HHD as a set of elliptic PDEs, solved using FFT for the base grid and iterative solvers for the refinement patches. Implemented in Fortran and parallelised with OpenMP, the code, named vortex, was tested on both idealised and complex cases, achieving errors below 1 per-mil in ideal cases and around 1% in more complex tests. This approach provides an accurate and adaptable tool for HHD in various simulations. The code is publicly available.

We investigated the nature of mass flows through the boundaries of cosmic voids to determine whether their velocity field is purely outflowing or more complex. Using a cosmological simulation tailored to describe voids and their surroundings, we extracted a sample of voids defined as large ellipsoids around expanding density minima. We then estimated gas mass fluxes through the void boundaries using a pseudo-Lagrangian approach. Surprisingly, we found that around 10% of the gas mass in voids at z = 0 has flowed in from denser regions, with some voids showing inflow fractions as high as 35%. Additionally, dark matter inflows persist within the inner regions of voids for extended periods. These findings suggest that, even in the largest voids, the velocity field is not purely outflowing. The presence of inflows, including gas from formerly denser areas, could significantly impact galaxy formation in underdense environments.

In this work, we critically examined how the assembly state of galaxies, groups and clusters (focusing on their DM haloes) can be best assessed at different redshifts. Using properties such as centre offset, virial ratio, mean radial velocity, sparsity, and ellipticity from a simulation, we studied how to combine these indicators to classify structures as relaxed, unrelaxed, or marginally relaxed, correlating them with merging and accretion activity. The classification was also validated against data from the CAMELS project, confirming that optimal thresholds and weights for these indicators vary strongly with redshift.

We studied the location and intensity of the outermost accretion shocks of galaxy clusters, aiming to characterise scaling relations could be used to measure cluster mass. Using a sample of galaxy clusters and groups from a cosmological simulation, we applied a shock-finding algorithm based on temperature jumps to identify shocks and devised a heuristic approach to characterise the accretion shocks of clusters and groups. We found that the clusters in our sample lie on a plane in the three-dimensional space defined by cluster mass, shock radius, and Mach number. Based on this relation, we propose that future observations of shock radii and intensities could allow for an indirect measurement of the cluster mass, with an error margin of around 30 percent at the 1-sigma level. This method would provide a new, independent avenue to measuring cluster masses and, thus, to constrain the ΛCDM model.