Thermal conductivity model for pure fluids and mixtures based on hidden scale invariance: Unifying gas, liquid, near-critical, and supercritical regions

The thermal conductivity of fluids plays an important role in heat transfer processes. This work develops a new thermal conductivity model based on hidden scale invariance. An exponential-linear superposition relationship between the residual thermal conductivity and reduced density is established. The model has three correlation parameters for each pure fluid. The physical significance of the three parameters is explained using intermolecular interaction forces. For mixtures, the model only requires a weighted average of the pure fluid parameters, without introducing additional interaction parameters. Combined with the dilute gas term and the critical term, the model overcomes the divergence in the dilute gas region and the critical enhancement effect. Experimental thermal conductivity data were collected for 123 pure fluids (63,661 data points) and 150 mixtures (13,687 data points). The model achieves an average absolute relative deviation (AARD) of 3.2% for pure fluids, with 93% of the experimental data falling within ±10%. For mixtures, the AARD is 4.3%, with 92% of the experimental data falling within ±10%. In addition, validation was performed using the widely applied models in REFPROP 10.0, and the AARD for both pure fluids and mixtures differ by less than 1% from those of the proposed models. It provides a reliable tool for fluid thermal conductivity prediction.