Metallic Glass · foundry-ml
dataset_metallicglass_rc
Experimental critical cooling rate measurements for 2,125 metallic glasses to support glass-forming ability modelling.
What dataset_metallicglass_rc is
dataset_metallicglass_rc provides experimental critical cooling rate data across 2,125 metallic glass systems. Developed by Benjamin T. Afflerbach, Carter Francis, Lane E. Schultz, and collaborators, the collection compiles empirical observations to aid the computational evaluation of glass-forming ability in alloys.
The critical cooling rate defines how swiftly a molten mixture must be cooled to bypass crystallisation and freeze into an amorphous state. Determining this threshold computationally addresses a longstanding bottleneck in materials science and engineering, where identifying viable bulk metallic glasses traditionally requires time-consuming experimental synthesis.
The dataset links alloy compositions and phase designations with log-transformed critical cooling rates, enabling both regression and classification approaches for alloy discovery and characterisation.
How it was made
The data is structured as a tabular collection stored in Parquet format with a single training partition. Each observation contains the material's chemical formula, normalized composition, material phase designation, and an indicator of the experiment type used to record the measurement.
The primary supervised target is logRc, representing the critical cooling rate on a logarithmic scale. Additional attributes include experimental quality assessments, prediction flags, and weighting factors.
Key facts
- Dataset size
- 2,125 metallic glasses
- Data format
- Tabular (Parquet)
- Primary target
- logRc (critical cooling rate, log scale)
- Primary inputs
- Chemical formula, composition, phase, experimental quality
- Supported tasks
- Tabular regression, tabular classification
- Publication year
- 2022
- Identifier
- DOI: 10.18126/rtsj-2e11
What you can use it for
- Cooling rate regression
- Train statistical and machine learning models to predict the log critical cooling rate directly from chemical compositions and elemental properties.
- Glass-forming ability classification
- Classify alloy formulations according to their phase designations and capacity to yield fully amorphous structures under specified cooling conditions.
- Quality-weighted materials screening
- Incorporate experiment quality indicators and sample weights into training objectives to evaluate how experimental technique influences model robustness.
Explore the data
The schema, real rows, full-text search and a SQL filter — read in your browser, nothing downloaded.
Explore
Read the real rows without downloading anything
Reading rows…
LiveRead from foundry-ml/dataset_metallicglass_rc at the moment you asked. Nothing is cached or stored — every row above came from that request.
How the columns are distributed
Over 2,125 rows of default/train
FORMULA
string_text- min
- 6
- median
- 11
- mean
- 12.09
- max
- 38
Prediction
int- min
- 0
- median
- 0
- mean
- 0
- max
- 0
composition
string_text- min
- 3
- median
- 13
- mean
- 13.94
- max
- 45
index
int- min
- 0
- median
- 1,571
- mean
- 1588.40
- max
- 8,891
logRc
float- min
- -2.05
- median
- 5.50
- mean
- 5.29
- max
- 8.44
phase
string_label- CR48%
- AC25%
- unknown16%
quality
string_label- meltspunCR48%
- meltspunAC25%
- exptdmax16%
weight
int- min
- 1
- median
- 1
- mean
- 1
- max
- 1
Strengths and limits
Good at
- Compiles 2,125 distinct experimental critical cooling rate records for metallic glasses.
- Supplies curated annotations including phase information, elemental composition, and experimental quality tags.
- Structured in standard tabular formats suitable for immediate use in numerical regression and classification pipelines.
Watch out for
- Provides only a single training split, requiring users to define their own cross-validation or holdout evaluation schemes.
- Contains tabular compositions and metadata but lacks direct time-temperature-transformation curves or raw thermal profiles.
Questions
Can I look inside dataset_metallicglass_rc without downloading it?
Yes. This page reads its schema, real rows and per-column statistics in the browser, with full-text search and a SQL filter. Nothing is downloaded to your machine.
What licence is dataset_metallicglass_rc under?
Its authors declare other. Read the terms on the original page before reusing the data.
What property does the target column logRc represent?
The logRc column represents the logarithm of the critical cooling rate, which measures the minimum quenching speed needed to form an amorphous metallic glass rather than a crystalline solid.
What features are included in the dataset?
The tabular features include the material formula, composition, phase designation, experimental quality descriptor, sample weight, entry index, and prediction values.
How large is the dataset?
The dataset contains 2,125 experimental entries of metallic glasses.
How to cite it
Afflerbach, Benjamin T. and Francis, Carter and Schultz, Lane E. and Spethson, Janine and Meschke, Vanessa and Strand, Elliot and Ward, Logan and Perepezko, John H. and Thoma, Dan and Voyles, Paul M. and Szlufarska, Izabela and Morgan, Dane, Machine Learning Prediction of the Critical Cooling Rate for Metallic Glasses from Expanded Datasets and Elemental Features, Materials Data Facility, 2022, https://doi.org/10.18126/rtsj-2e11
Written from dataset_metallicglass_rc's own dataset card. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.