Numerical simulations of highly porous dust aggregates in the low-velocity collision regime Implementation and calibration of a smooth particle hydrodynamics code

Context. A highly favoured mechanism of planetesimal formation is collisional growth. Single dust grains hit each other with relative velocities produced by gas flows in the protoplanetary disc. They stick together with van der Waals forces and form fluffy aggregates up to a centimetre size. The mechanism responsible for any additional growth is unclear since the outcome of aggregate collisions in the relevant velocity and size regime cannot be investigated in the laboratory under protoplanetary disc conditions. Realistic statistics for dust aggregate collisions beyond decimetre size are required to obtain a deeper understanding of planetary growth. Aims. By combining experimental and numerical efforts, we wish to calibrate and validate a computer program capable of accurately simulating the macroscopic behaviour of highly porous dust aggregates. After testing its numerical limitations thoroughly, we check the program especially for a realistic reproduction of the compaction, bouncing, and fragmentation behaviour. This demonstrates the validity of our code, which will be utilised to simulate dust aggregate collisions and accurately determine the fragmentation statistics in future work. Methods. We adopt the smooth particle hydrodynamics (SPH) numerical scheme with extensions to the simulation of solid bodies and a modified version of the Sirono porosity model. Experimentally measured macroscopic material properties of SiO2 dust are implemented. By simulating three different setups, we calibrate and test for the compressive strength relation (compaction experiment) and the bulk modulus (bouncing and fragmentation experiments). Data from experiments and simulations will be compared directly. Results. SPH has already proven to be a suitable tool for simulating collisions at rather high velocities. In this work, we demonstrate that its area of application may be beyond low-velocity experiments and collisions. It can also be used to simulate the behaviour of highly porous objects in this velocity regime to very high accuracy. A correct reproduction of density structures in the compaction experiment, of the coefficient of restitution in the bouncing experiment, and of the fragment mass distribution in the fragmentation experiment illustrate the validity and consistency of our code for the simulation of the elastic and plastic properties of the simulated dust aggregates. The result of this calibration process is an SPH code that can be utilised to investigate the collisional outcome of porous dust in the low-velocity regime.