Trihybrid Nanofluid EMHD Flow Simulation Data for ANN Training

S. B. Parida published a dataset of 200 data points on 2026-06-02, derived from numerical simulations of electromagnetohydrodynamic (EMHD) flow. The data supports the training and validation of an Artificial Neural Network (ANN) model for predicting skin friction, Nusselt number, and Sherwood number in a trihybrid nanofluid system. The dataset is small at 50.3 KB and includes results from simulations considering variable viscosity, viscous dissipation, and convective boundary conditions.

Use Cases

• Training ANN models to predict skin friction coefficients based on simulation parameters.
• Validating ANN performance for predicting Nusselt number (heat transfer) in trihybrid nanofluid flows.
• Testing ANN architectures for predicting Sherwood number (mass transfer) under EMHD conditions.
• Benchmarking numerical simulation results against ANN predictions for entropy generation analysis.

Strengths

• Dataset includes 200 data points with a defined 70/15/15 split for training, testing, and validation.
• ANN model performance is documented with specific metrics: for skin friction, R-values reached 0.98115 (training) and 0.97209 (overall).
• Convergence details are provided, such as skin friction model converging at 99 epochs with an MSE of 0.0275.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small (50.3 KB), indicating limited scope for large-scale model training.

Provenance

Source

figshare, author S. B. Parida.

Collection Method

Data generated from numerical solutions of transformed ODEs using the fourth-order Runge-Kutta method with shooting and MATLAB's bvp4c solver.

Freshness

Last updated 2026-06-02 06:58:44; freshness should be verified.