ndemo commited on
Commit
25e44ab
·
verified ·
1 Parent(s): cfe21f6

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +82 -0
README.md CHANGED
@@ -49,3 +49,85 @@ configs:
49
  - split: default
50
  path: snapshots/default-*
51
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
49
  - split: default
50
  path: snapshots/default-*
51
  ---
52
+ # Thermal Block Dataset
53
+
54
+ ## Dataset Description
55
+
56
+ This dataset contains thermal diffusion simulations on a 2D block geometry with varying thermal conductivity parameters.
57
+
58
+ ### Dataset Summary
59
+
60
+ The Thermal Block dataset provides numerical simulations of heat transfer in a 2D block with parametrized thermal conductivity. The dataset is useful for reduced-order modeling, surrogate modeling, and physics-informed machine learning applications in thermal analysis.
61
+
62
+ ## Dataset Structure
63
+
64
+ ### Data Instances
65
+
66
+ The dataset consists of three configurations:
67
+
68
+ - **geometry**: Mesh information (nodes and connectivity)
69
+ - **snapshots**: Temperature field solutions
70
+ - **parameters**: Thermal conductivity values for each simulation
71
+
72
+ ### Data Fields
73
+
74
+ #### Geometry Configuration
75
+ - `node_coordinates_x`: Sequence of x-coordinates of mesh nodes (float64)
76
+ - `node_coordinates_y`: Sequence of y-coordinates of mesh nodes (float64)
77
+ - `connectivity`: Sequence of element connectivity (triangular elements, int32)
78
+
79
+ #### Snapshots Configuration
80
+ - `temperature`: Temperature field at each node (float64)
81
+
82
+ #### Parameters Configuration
83
+ - `conductivity`: Thermal conductivity parameter for each simulation (float64)
84
+
85
+ ### Data Splits
86
+
87
+ - `default`: Contains all simulations
88
+
89
+ ## Dataset Creation
90
+
91
+ ### Source Data
92
+
93
+ The dataset was generated using finite element simulations of the heat equation with varying thermal conductivity parameters.
94
+
95
+ ### Preprocessing
96
+
97
+ Solutions are stored as 1D arrays corresponding to the nodal values on the mesh.
98
+
99
+ ## Usage
100
+
101
+ ```python
102
+ from datasets import load_dataset
103
+ import numpy as np
104
+ import matplotlib.pyplot as plt
105
+ import matplotlib.tri as mtri
106
+
107
+ # Load geometry
108
+ ds_geom = load_dataset("SISSAmathLab/thermal-block", name="geometry")
109
+
110
+ # Load snapshots
111
+ ds_data = load_dataset("SISSAmathLab/thermal-block", name="snapshots")
112
+
113
+ # Load parameters
114
+ ds_params = load_dataset("SISSAmathLab/thermal-block", name="parameters")
115
+
116
+ # Visualize temperature distribution for first simulation
117
+ pts_x = np.asarray(ds_geom['default']['node_coordinates_x']).flatten()
118
+ pts_y = np.asarray(ds_geom['default']['node_coordinates_y']).flatten()
119
+ connectivity = ds_geom['default']['connectivity'][0]
120
+ temperature = ds_data['default']['temperature'][0]
121
+
122
+ triang = mtri.Triangulation(pts_x, pts_y, connectivity)
123
+ plt.tripcolor(triang, temperature)
124
+ plt.colorbar(label='Temperature')
125
+ plt.title('Thermal Block - Temperature Distribution')
126
+ plt.xlabel('x')
127
+ plt.ylabel('y')
128
+ plt.show()
129
+ ```
130
+
131
+ ## Contact
132
+
133
+ For questions or issues, please contact SISSA mathLab.