Datasets:

Formats:
parquet
License:
Dataset Viewer
Auto-converted to Parquet Duplicate
Search is not available for this dataset
optimal_design
listlengths
80k
80k
inlet_velocity
float64
-0.1
-0.03
max_power_dissipation
float64
47.7
75
volfrac
float64
0.25
0.7
mean_temperature
float32
7.26
71.3
power_dissipation
float32
46.3
75.4
[0.0003150000120513141,0.00013840000610798597,0.0000583200016990304,0.000015499999790336005,6.967999(...TRUNCATED)
-0.0705
50.1
0.67
9.733
50.261902
[0.0004594000056385994,0.00017010000010486692,0.00005848000000696629,0.000017210000805789605,6.76799(...TRUNCATED)
-0.0901
60.5
0.64
7.92828
60.225498
[7.675999995626626e-7,1.0109999948326731e-6,1.690999965830997e-6,2.998999889314291e-6,4.730999989988(...TRUNCATED)
-0.0594
59.2
0.34
15.7451
59.284
[1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0(...TRUNCATED)
-0.0568
66.3
0.63
12.0503
65.0952
[1.4460000556937302e-6,8.247999971899844e-7,9.455000054003904e-7,1.6069999446699512e-6,3.68099995284(...TRUNCATED)
-0.0613
59.1
0.37
13.6411
59.349098
[0.000020900000890833326,1.3139999737177277e-6,2.767000069070491e-7,2.0119999533108057e-7,1.90299999(...TRUNCATED)
-0.0458
52.7
0.68
14.8519
52.149601
[0.000010660000043571927,6.210999799804995e-6,5.1649999477376696e-6,4.759000148624182e-6,6.847999884(...TRUNCATED)
-0.0647
63.6
0.55
10.9396
63.664902
[8.371999683731701e-6,7.97200027591316e-6,9.053999747266062e-6,0.000012460000107239466,0.00002207999(...TRUNCATED)
-0.092
59.1
0.43
10.6628
59.2715
[0.00003147000097669661,0.00003212999945390038,0.00004484999954001978,0.00003470999945420772,0.00001(...TRUNCATED)
-0.0728
70.7
0.44
10.6933
70.776001
[0.00001083999995898921,7.103999905666569e-6,6.37399989500409e-6,5.723999947804259e-6,6.700000085402(...TRUNCATED)
-0.0776
63.5
0.44
10.5456
63.519501
End of preview. Expand in Data Studio

MTO2D v0

EngiBench-compatible MTO2D dataset published as IDEALLab/mto_2d_v0. It contains 5,666 converged (400, 200) half-domain designs with train/val/test splits of 4,249/283/1,134.

Objectives: mean_temperature and power_dissipation are historical source-optimization labels logged before the final MMA/Heaviside update. The stored app/200/gamma design is the subsequent post-update density field.

For the EngiBench problem, a frozen simulator uses the following physics: q=0.01, alphaMax=5025200, Heaviside=59.8, with design updates disabled.

optimal_design is a flattened float list of length 80,000, reshaping to (400, 200) and symmetrical about the vertical axis so as to yield a (400, 400) heat sink.

Citation

@article{drake2026quantize,
  title={To Quantize or Not to Quantize: Effects on Generative Models for Topology Optimization Problems},
  author={Drake, Arthur and Chen, Qiuyi and Wang, Jun and Nejat, Ardalan and Guest, James K and Fuge, Mark},
  journal={Journal of Mechanical Design},
  volume={148},
  number={10},
  pages={101704},
  year={2026},
  publisher={American Society of Mechanical Engineers}
}
Downloads last month
44