Fluid Acceleration in Heavy-Ion Collisions

For both models, the spatial range of the simulation grid is −10 fm≤x,y,z≤10 fm with a spatial grid step (resolution) of DX=DY=DZ=0.5 fm. The temporal evolution covers 0≤t≤10 fm/c for the AMPT model, and 0≤t≤80 fm/c for the UrQMD model. The sampling time step is 0.1 fm/c for both models.

The central difference method uses a time derivative step of 2×dt=0.1 fm/c (with dt=0.05 fm/c) for AMPT, and 2×dt=0.2 fm/c (with dt=0.1 fm/c) for UrQMD. For the 2D transverse plane distribution data, the grid size is extended by 2 empty padding bins on each boundary, resulting in a 45×45 grid size covering a physical range of −11.5 fm≤x,y≤11.5 fm with the same spatial resolution of 0.5 fm.The dataset is organized into four folders matching the figures in the published paper.

Figure 1 contains the 2D transverse spatial distribution of acceleration in a three-column format (X-coordinate, Y-coordinate, and value). The coordinates are in fm, and the acceleration value is in GeV/fm. Figures 2 and 3 represent curves of impact-parameter dependence and full-space time evolution of average acceleration.

The first column of these files denotes the independent variable (b in fm or t in fm/c), and the subsequent columns denote the acceleration in MeV for different collision energies. Figure 4 lists the average acceleration components (ax,ay,az) in MeV in the columns time(fm/c) ax ay az. Any grid cells that fell below the physical energy density cutoff are assigned a value of 0.0.The numerical precision of all datasets is in double-precision floating-point format, presented with 6 to 8 decimal places in the .dat files.

As the dataset represents statistical averages over 1000 (AMPT) and 200 (UrQMD) events, the statistical errors arising from thermal and event-by-event fluctuations are well-controlled. All files are saved in the standard plain-text .dat format, which can be easily parsed using standard tools such as Python, MATLAB or ROOT.