|
| 1 | +# Apollo 16 booster impact |
| 2 | + |
| 3 | +--- |
| 4 | + |
| 5 | + |
| 6 | + |
| 7 | +This is a simple example for a Moon simulation. It takes the lunar topography and sets up a model |
| 8 | +around the Apollo 16 booster impact and the seismic station installed by Apollo 12. |
| 9 | + |
| 10 | +The Apollo 16 S-IVB booster impact (Onodera et al. 2022; Wagner et al. 2017) details are: |
| 11 | +* time of impact approximately on April 19, 1972, 21:02:04 (UTC) |
| 12 | +* location is at lat = 1.9210, lon = 335.3770 (deg) |
| 13 | +* impact angle specified by angle from horizon: 79, heading angle (NE): 255.3 (deg) |
| 14 | +* relatively low impact speed v ~ 2.6 km/s, and assumed booster mass (stage at impact) m ~ 13,973 kg |
| 15 | +* momentum p = m * v ~ 3.6 * 10^7 [kg m/s], with an estimated deceleration time dt ~ 0.1 s, |
| 16 | + leads to an average force F = p / dt ~ 3.6 * 10^8 [N] |
| 17 | + |
| 18 | + |
| 19 | +Moon stations: |
| 20 | + ``` |
| 21 | + S12 XA -3.01084 -23.42456 0.0 0.0 # Apollo 12 |
| 22 | + S14 XA -3.64450 -17.47753 0.0 0.0 |
| 23 | + S15 XA 26.13407 3.62981 0.0 0.0 |
| 24 | + S16 XA -8.97577 15.49649 0.0 0.0 |
| 25 | + ``` |
| 26 | + |
| 27 | +For meshing, we use a Moho interface at an estimated 30.8 km depth (at Apollo 12 landing site) and a 3-layer model. |
| 28 | +The top, middle and bottom layers have regolith, crustal and mantle rock properties respectively. |
| 29 | +The values are taken from Garcia et al. (2011) VPREMOON model. |
| 30 | + |
| 31 | +---- |
| 32 | + |
| 33 | +## Setup |
| 34 | + |
| 35 | +The setup of the model is meant for a low-resolution movie simulation that runs on 8 CPU-cores. |
| 36 | +The mesh is build for the region around the booster impact and the Apollo 12 seismic station, |
| 37 | +an area from longitude -30 to -20 degrees, latitude from -5 to 5 degrees. |
| 38 | + |
| 39 | +The setup is done by the following steps: |
| 40 | + |
| 41 | +* **Step 1:** Mesh setup |
| 42 | + |
| 43 | + We will use the in-house mesher `xmeshfem3D` to create our simulation mesh. The main files in `DATA/meshfem3D_files/Mesh_Par_file` |
| 44 | + and `interfaces.dat` are setup such that we will have 3 layers, where the Moho surface at 30.8 km depth is added manually by |
| 45 | + creating a flat interface file `interface.moho.dat`. |
| 46 | + |
| 47 | + To facilitate the meshing with doubling-layers, we also add an additional intermediate flat interface at 8 km depth from where the mesh |
| 48 | + will start the stretching to account for the topography. This avoids issues with elements having negative Jacobians due to the stretching |
| 49 | + and doubling of the mesh element sizes. |
| 50 | + |
| 51 | + In `interfaces.dat`, we also added two more entries for the surfaces `ptopo.xyz.*.dat` that will be created and modified accordingly in the next step. |
| 52 | + |
| 53 | + |
| 54 | +* **Step 2:** Topography |
| 55 | + |
| 56 | + To create the mesh topographic surface of the Moon, we do: |
| 57 | + ``` |
| 58 | + > cd EXAMPLES/real_world/Moon_Apollo_impact |
| 59 | + > ln -s ../../../utils/scripts/run_get_simulation_topography.py |
| 60 | + ``` |
| 61 | + and then |
| 62 | + ``` |
| 63 | + > ./run_get_simulation_topography.py -30 -5 -20 5 --SRTM=moon --toposhift=1000.0 --toposcale=0.9 |
| 64 | + ``` |
| 65 | + |
| 66 | + This script creates the files needed in a subfolder `topo_data/`, in particular the topography surface (`ptopo.xyz.1.dat`) and |
| 67 | + a down-shifted surface (`ptopo.xyz.2.dat`, by approximately 1 km and a slight down-scaling of the topography by factor `0.9`) |
| 68 | + to defined our regolith elements within this layer. |
| 69 | + |
| 70 | + The script also modified the entries in the `interfaces.dat` to read the correct number of entries and increments. |
| 71 | + It also modifies `Mesh_Par_file` to set the correct mesh lat/lon-dimensions and lunar projection zone number (as the `UTM_PROJECTION_ZONE`). |
| 72 | + |
| 73 | + > [!NOTE] |
| 74 | + > For Moon simulations, instead of the Universal Transverse Mercator (UTM) projection valid for Earth, the mesher will use a Lunar Transverse Mercator (LTM) projection (for latitudes between [-82,82] degrees) or a Lunar Polar Stereographic (LPS) at North/South pole regions. To enable these lunar projections, use a model name starting with `moon_***` (see `moon_default` in `Par_file`; or use `moon_tomo` for combining it with a tomography model). |
| 75 | + > The zone numbers are positive for the Northern hemisphere and negative for Southern hemisphere. LTM uses zone numbers in the range +/- [1,45], LPS uses 46 for North pole and -46 for South pole regions. The script `run_get_simulation_topography.py` will output the corresponding zone number for the specified area (using its midpoint to determine the zone). |
| 76 | +
|
| 77 | + |
| 78 | + |
| 79 | +* **Step 3:** source and station |
| 80 | + |
| 81 | + The station "S12" added in `DATA/STATIONS` is the Apollo 12 seismic station at location lat/lon = (-3.01084, -23.42456) (deg). |
| 82 | + For the Apollo 16 S-IVB booster impact, we use a point force `DATA/FORCESOLUTION` at location lat/lon = (1.9210, 335.3770) (deg). |
| 83 | + The source time function is a Gaussian with a force factor and direction estimated by the informations given above. |
| 84 | + |
| 85 | + |
| 86 | +For convenience, this example folder has all these required setup files to run directly the wave simulation. |
| 87 | + |
| 88 | +--- |
| 89 | + |
| 90 | +## Scattering perturbations |
| 91 | + |
| 92 | +For this moon model, we will add random 3D scattering perturbations with a von Karman distribution in the regolith and crustal layers. The following entries can be added in the `DATA/Par_file` to add such perturbations: |
| 93 | +``` |
| 94 | +## scattering |
| 95 | +# adds scattering perturbations to velocity model |
| 96 | +SCATTERING_PERTURBATIONS = .true. |
| 97 | +# perturbation strength |
| 98 | +SCATTERING_STRENGTH = 0.4d0 |
| 99 | +# correlation factor |
| 100 | +SCATTERING_CORRELATION = 5.d0 |
| 101 | +# list of material ids to apply scattering perturbations (0 == for all materials) |
| 102 | +SCATTERING_MATERIAL_IDS = 2,3 |
| 103 | +``` |
| 104 | +Here we chose some arbitrary values just to visualize its effect. The settings will create perturbations with a strength of 40% and a von Karman correlation factor 5.0 that leads to an approximated correlation length of ~664 m (see output in `OUTPUT_FILES/output_generate_databases.txt`). The material IDs "2,3" are depicting the crustal and regolith material layers defined in the mesh by `Mesh_Par_file`. |
| 105 | + |
| 106 | + |
| 107 | +---- |
| 108 | + |
| 109 | +## Wave simulation: |
| 110 | + |
| 111 | + To run our simulation, just type: |
| 112 | + ``` |
| 113 | + > ./run_this_example.sh |
| 114 | + ``` |
| 115 | + |
| 116 | + The simulation is meant as a movie simulation. To suppress numerical noise for a cleaner movie visualization, we enlarged the source half-duration |
| 117 | + by `HDUR_MOVIE = 8.0` in the `Par_file`. |
| 118 | + |
| 119 | + The solver will create binary movie files in `OUTPUT_FILES/` folder. To convert them to VTK-files that can be visualize e.g. with Paraview, |
| 120 | + we run the script: |
| 121 | + ``` |
| 122 | + > ./xcreate_movie_files.sh |
| 123 | + ``` |
| 124 | + |
| 125 | + |
| 126 | + |
| 127 | + |
0 commit comments