The standard LAMMPS installation from package managers (such as apt) or pre-compiled binaries may not include all the features required by Trybo. Specifically, Trybo requires LAMMPS to be compiled with GPU acceleration and several additional packages.
To ensure compatibility, we provide a build script (toolchain/build_lammps_gpu.sh) that compiles LAMMPS with all necessary features. This script:
- Will not modify or replace any existing LAMMPS installation on your system
- Creates a custom LAMMPS executable in your current working directory
- Configures all required packages and optimizations for Trybo
You must use this custom executable when running Trybo simulations to avoid compatibility errors and to benefit from GPU acceleration.
The compute cluster/atom command identifies connected groups of atoms based on a cutoff distance, assinging a unique cluster id to each atom. The results are saved periodically in debris_clusters.dump. This enables the detection and analysis of material fragments that may detach from the main bodies (nanoparticle or planes) during friction.
This analysis specifically identifies and tracks atoms with particularly high kinetic energy ("hotspots") using compute ke_hotspots and a threshold (variable hot_threshold, variable is_hotspot). The script counts the total number of hotspots over time (compute hotspots_count, fix hotspots_monitor) and dumps per-atom kinetic energy and hotspot_status (dump hotspots).
This analysis focuses on the distribution and evolution of kinetic, potential and total energy per atom for the "mobile" group of atoms (nanoparticle + flexible lower plane atoms). The compute ke/atom and compute pe/atom commands calcualte these per-atom energies, which are then summed to get total energy per atom (variable total_energy). This data is periodically saved via dump energy_dump.
Adhesives wear involves the transfer of material between contacting surfaces. The script uses dynamic groups and atom counts (v_transferred_to_lower_group, v_transferred_to_upper_group) to estimate the net transfer of atoms from the initial groups into regions defined as the lower and upper planes. This data is saved over time in material_transfer.txt and included in the thermo output.
Polyhedral Templat
8000
e Matching (PTM), calculated using compute ptm/atom and dumped per atom in ptm.dump, is an alternative method to CNA for identifying local crystal structures (FCC, HCP, BCC, surface, interface, unknown). It provides a detailed characterization of the local atomic envirionment and is particularly useful for analyzing structural changes during deformation and wear.
Beyond the hotspot analysis, the script explicitly calculates the temperature of the mobile_group (c_temp_mobile) and specifically the temperature of the nanoparticle_group (v_nanoparticle_temperature) based on atomic velocities. These temperatures, along with the overal system temperature (temp), are available in the thermo output, and the nanoparticle temperature is saved in wear_data.txt. This allows for studying the thermal response of the system and the nanoparticle due to frictional heating and thermostatting.
This analysis calcualtes the von Mises stress for each atom using compute vonmises and variable atoms_stresss. This scalar value is often used as an indicator of plastic deformation and yielding. The per-atom von Mises stress is dumped via dump vonmises_dump, and the average and maximum von Mises stress values are calculated over time (compute vonmises_average, compute vonmises_max) and saved in wear_data.txt.
Focuses on the forces exchanged between the nanoparticle and the upper plane wich are fundamental to friction. The script calculates the normal force (v_fz_nanoparticle_upper_plane) and the tangential friction force (v_fx_nanoparticle_upper_plane) between the nanoparticle group and the upper plane group. It algo calculates the overall friction force on the upper plane (v_fx_upper_plane) and the coefficient of friction (v_coefficient_friction, v_coefficient_friction_capped). These values are available in thermo output and the coefficient of friction, is saved over time in friction_data.txt
A direct measure of mechanical wear is the loss of material from the nanoparticle. The script tracks the number of atoms currently in the nanoparticle_group using variable contact_count. This count is saved over time in contact_count.txt and wear_data.txt, and is included in the thermo output.
This analysis utilizes the compute cna/atom to assign a value to each atom indicating its local structural environment (e.g., perfect FCC, perfect HCP surface, or various defect types). The results are saved periodically for all atoms using dump cna_dump. The structural changes are key indicators of material degradation and deformation mechanisms that ocurr during wear processes.
The analysis is based on the compute coord/atom command, which calculates the number of nearest neighbors for each atom within a specified cutoff distance. The script dumps this per-atom coordination number (c_coord) via dump coord_dump and also calculates and saves the average, minimum and maximum coordination numbers over time using compute coord_avg, compute coord_min, compute coord_max, and fix coord_stats.
The centro-symmetric parameter (c_center_symmetric), calculated using compute centro/atom, is a sensitive indicator of local structural deffects and disorder in crystalline material like FCC. The per-atom values are dumped in center_symmetric.dump, allowing for spatial analysis. The script also calculates and saves the average (c_center_symmetric_average), maximum (c_center_symmetirc_max), and a calculated defect percentage (v_defect_percent) based on this parameter over time in wear_data.txt and the thermo output.
The Radial Distribution Function (c_radial_distribution_function[*]), calculated using compute rdf and saved over time in radial_distribution_function_output.txt, describes the average atomic density as a function of distance from an atom. It provides information about the overall structural order and average interatomic distances in the system.
The thermo output provides a snapshot of global system properites at regular intervals. This includes the simulation time step, total simulation time, total number of atoms, system pressure (press), total energy (etotal), potential energy (pe) and kinetic energy (ke). It also tracks the commanded x-position of the upper plane (v_x_pos) and its average z-position (v_z_upper_plane).