Skip to content

SoumyMittal/nanofluid-thermal-conductivity-ml-dataset

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Unified Thermal Conductivity Dataset of Nanofluids

Dataset Preview

Nanoparticle BaseFluid Temperature (°C) Particle Size (nm) TC (W/mK)
Al₂O₃ Water 25 30 0.67
CuO EG 40 50 0.82

Overview

The Unified Thermal Conductivity Dataset of Nanofluids is a machine-learning-ready dataset containing 987 experimental records collected and consolidated from multiple peer-reviewed nanofluid studies.

This dataset was developed to support research in:

  • Machine Learning for Material Informatics
  • Thermal Conductivity Prediction
  • Data-Driven Thermal Engineering
  • Scientific Computing
  • Feature Engineering for Engineering Applications

The dataset combines experimental measurements and engineered features to facilitate predictive modeling and analysis of nanofluid thermal conductivity.


Dataset Statistics

Attribute Value
Total Records 987
Features 12
Domain Thermal Engineering / Material Informatics
Data Source Peer-Reviewed Research Studies
Task Type Regression
Format CSV

Features

Feature Description
Nanoparticle Type of nanoparticle used
BaseFluid Base fluid used in the nanofluid
Temperature (°C) Experimental temperature
Particle Size (nm) Average nanoparticle size
Particle Volume Fraction (%) Concentration of nanoparticles
TC (W/mK) Thermal conductivity of nanofluid (Target Variable)
Thermal Conductivity of Particle (kp) Thermal conductivity of nanoparticle
Thermal Conductivity of Liquid (km) Thermal conductivity of base fluid
kp/km ratio Ratio of particle to fluid conductivity
Temp_Volume Temperature-volume interaction feature
Inverse_Particle_Size Reciprocal particle size feature
Brownian_Feature Engineered feature representing Brownian motion effects

Applications

This dataset can be used for:

  • Thermal Conductivity Prediction
  • Machine Learning Regression Models
  • Feature Importance Analysis
  • Material Informatics Research
  • Engineering Data Analytics
  • Scientific Machine Learning
  • Comparative Model Evaluation

Potential Machine Learning Models

Researchers and students can use this dataset with:

  • Linear Regression
  • Random Forest Regressor
  • XGBoost
  • Gradient Boosting
  • Support Vector Regression (SVR)
  • Artificial Neural Networks
  • Ensemble Learning Methods

Feature Engineering

To improve machine learning performance, several domain-informed features were engineered:

  • kp/km Ratio to represent relative conductivity effects
  • Temperature-Volume Interaction to capture coupled thermal behavior
  • Inverse Particle Size to model nanoscale size-dependent effects
  • Brownian Feature to represent particle motion influence on heat transfer

These features were designed using engineering knowledge from nanofluid heat-transfer literature.


Research Motivation

Nanofluids have gained significant attention due to their enhanced thermal properties and applications in heat transfer systems.

The objective of this dataset is to provide a standardized and machine-learning-ready benchmark for researchers interested in applying artificial intelligence and data-driven techniques to thermal conductivity prediction.


Citation

If you use this dataset in academic work, research projects, or publications, please cite this repository and acknowledge the original experimental studies from which the data was compiled.


Author

Soumy Mittal

B.Tech Artificial Intelligence & Data Science Amrita Vishwa Vidyapeetham

Research Interests

  • Machine Learning
  • Material Informatics
  • Scientific Computing
  • Computational Modelling
  • Data-Driven Engineering
  • Intelligent Systems

License

This dataset is released under the MIT License.

About

A curated dataset for machine learning applications in nanofluid thermal conductivity prediction.

Topics

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors