Alloy Properties · foundry-ml
dataset_mg_alloy
A tabular dataset of 365 biomedical magnesium alloys linking composition and processing to mechanical properties.
What dataset_mg_alloy is
dataset_mg_alloy is a curated tabular dataset designed for machine learning modelling of lightweight structural and biomedical magnesium alloys. Developed by Haobing Hou, Jianfeng Wang, Li Ye, Shijie Zhu, Liguo Wang, and Shaokang Guan, it provides experimental records detailing elemental compositions, thermal and mechanical processing histories, and resulting mechanical performance metrics.
Magnesium alloys are of significant interest in biomedical engineering and structural design due to their low density and biocompatibility. However, predicting how specific alloying elements and multi-stage thermo-mechanical treatments impact strength and ductility remains challenging. This dataset addresses that problem by cataloguing processing parameters alongside elemental fractions across 365 alloy samples.
The dataset enables computational materials scientists to develop tabular regression and classification models to predict tensile properties, aiding in the discovery and optimisation of biomedical magnesium alloys without requiring exhaustive physical synthesis.
How it was made
The dataset is structured into tabular format comprising 365 magnesium alloy instances with single split structure (train). For each alloy, input features include elemental mass fractions in weight percent (Mg, Zn, Y, Zr, Nd, and Gd) alongside processing parameters across various manufacturing steps.
Processing inputs record the thermal and mechanical parameters of alloy fabrication: solution treatment temperature and duration, homogenization temperature and duration, extrusion temperature and extrusion ratio, and artificial aging temperature and duration. The primary target properties are ultimate tensile strength (UTS in MPa), yield strength (YS in MPa), and elongation (EL in percentage).
Key facts
- Number of samples
- 365
- Data format
- Tabular (Parquet)
- Input features
- 14 features (6 elemental compositions, 8 thermo-mechanical processing parameters)
- Target properties
- Ultimate tensile strength (UTS), yield strength (YS), elongation (EL)
- Primary elements
- Mg, Zn, Y, Zr, Nd, Gd
- Publication year
- 2023
- DOI
- 10.18126/myj4-0h48
What you can use it for
- Predicting tensile strength
- Train tabular regression models to forecast ultimate tensile strength and yield strength based on elemental compositions and thermo-mechanical processing conditions.
- Predicting alloy ductility
- Model elongation percentages to determine how specific processing schedules and alloying additions influence alloy ductility.
- Optimising processing parameters
- Evaluate the sensitivity of mechanical outcomes to extrusion temperature, extrusion ratio, solution treatment, and aging schedules for magnesium alloy design.
Explore the data
The schema, real rows 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_mg_alloy at the moment you asked. Nothing is cached or stored — every row above came from that request.
How the columns are distributed
Over 365 rows of default/train
EL(%)
float- min
- 0.50
- median
- 10.30
- mean
- 11.43
- max
- 37.59
Gd(wt.%)
float- min
- 0
- median
- 0
- mean
- 1.43
- max
- 20
Mg(wt.%)
float- min
- 75.20
- median
- 92.42
- mean
- 92.44
- max
- 99.27
Nd(wt.%)
float- min
- 0
- median
- 0
- mean
- 0.58
- max
- 4
UTS(MPa)
float- min
- 75
- median
- 287.20
- mean
- 287.49
- max
- 483
Y(wt.%)
float- min
- 0
- median
- 0.61
- mean
- 1.69
- max
- 18.14
YS(MPa)
float- min
- 20.90
- median
- 204
- mean
- 222.16
- max
- 425
Zn(wt.%)
float- min
- 0
- median
- 3.74
- mean
- 3.53
- max
- 15.17
Zr(wt.%)
float- min
- 0
- median
- 0.47
- mean
- 0.34
- max
- 2
aging temperature(°Ê)
int- min
- 0
- median
- 0
- mean
- 53.44
- max
- 250
aging time(h)
float- min
- 0
- median
- 0
- mean
- 9.57
- max
- 126
extrusion ratio
float- min
- 0
- median
- 10
- mean
- 10.37
- max
- 42
extrusion temperature(°Ê)
int- min
- 0
- median
- 300
- mean
- 226.47
- max
- 500
homogenization temperature(°Ê)
int- min
- 0
- median
- 0
- mean
- 129.01
- max
- 560
homogenization time(h)
int- min
- 0
- median
- 0
- mean
- 5.68
- max
- 48
solution temperature(°Ê)
int- min
- 0
- median
- 0
- mean
- 150.27
- max
- 550
solution time(h)
float- min
- 0
- median
- 0
- mean
- 3.28
- max
- 25
Strengths and limits
Good at
- Combines both chemical composition and granular thermo-mechanical processing variables across multiple manufacturing stages.
- Provides simultaneous coverage of primary mechanical engineering properties, including ultimate tensile strength, yield strength, and elongation.
- Pre-processed and structured into standard tabular formats compatible with standard machine learning libraries.
Watch out for
- Contains a relatively modest sample size of 365 alloy instances, which limits the training of complex, highly parameterized architectures.
- Restricted to specific alloying additions (Zn, Y, Zr, Nd, and Gd), offering no coverage of other common magnesium alloying systems.
- Does not include explicit predefined validation or test splits, containing only a single training split in the repository.
Questions
Can I look inside dataset_mg_alloy without downloading it?
Yes. This page reads its schema, real rows and per-column statistics in the browser and a SQL filter. Nothing is downloaded to your machine.
What licence is dataset_mg_alloy under?
Its authors declare other. Read the terms on the original page before reusing the data.
What elements are included in the alloy compositions?
The dataset records weight percentages for magnesium (Mg), zinc (Zn), yttrium (Y), zirconium (Zr), neodymium (Nd), and gadolinium (Gd).
Which processing stages are captured in the feature set?
The dataset tracks solution temperature and time, homogenization temperature and time, extrusion temperature and ratio, and aging temperature and time.
What target variables can be predicted?
There are three primary targets: ultimate tensile strength (UTS) in MPa, yield strength (YS) in MPa, and elongation (EL) as a percentage.
How many alloy samples are available in the dataset?
The dataset contains records for 365 distinct magnesium alloy specimens.
How to cite it
Hou, Haobing and Wang, Jianfeng and Ye, Li and Zhu, Shijie and Wang, Liguo and Guan, Shaokang. Prediction of mechanical properties of biomedical magnesium alloys based on ensemble machine learning. Materials Data Facility (2023). https://doi.org/10.18126/myj4-0h48
Written from dataset_mg_alloy'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.