Shear Viscosity Models for 952 Pure Substances Based on 100,988 Data Points

100,988 experimental data points underpin viscosity models for 952 pure substances, correlating shear viscosity with residual entropy. Timo Klenk published this dataset on figshare in May 2026, proposing a revised ansatz function and a conservative outlier removal approach. The models achieve an average mean absolute relative deviation of 3.1%.

Use Cases

• Predicting gas-phase viscosities of real substances based on the proposed 'reference' viscosity model
• Extrapolating viscosity behavior under low-density conditions using the revised ansatz function
• Benchmarking new entropy-scaling models against a large collection of experimental data points
• Identifying and removing potential outliers in experimental viscosity measurements using the described conservative approach

Strengths

• Based on 100,988 experimental data points, which exceeds earlier parametrization studies
• Covers 952 pure substances, providing a broad scope of chemical compounds
• Model achieves an average mean absolute relative deviation of 3.1%
• Parametrization scheme is designed for robust extrapolation capabilities beyond the experimental data range

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• Experimental data are described as scarce and unevenly distributed for many substances, leading to data gaps

Provenance

Source

figshare

Collection Method

Parametric modeling of experimental viscosity data using a revised entropy-scaling approach.

Freshness

Last updated 2026-05-18 19:05:28