CFD4DC: Automated CFD framework for heat transfer analysis of data centers

Computational Fluid Dynamics (CFD) is increasingly applied to the thermal analysis of air-cooled data centers, particularly for retrofitting and cooling optimization. While CFD provides detailed insights into airflow and heat transfer, its use remains time-consuming and labor-intensive. This study introduces CFD4DC (Computational Fluid Dynamics for Data Centers), an in-house framework that automates the preparation and execution of CFD simulations for data center thermal management. The automation substantially reduces and simplifies required input data, offering advantages for both scientific studies and industrial applications. The framework is validated in a two-rack micro data center against experimental measurements, achieving a mean absolute error of 0.63 °C in the inlet temperatures of servers and air conditioners. These results demonstrate that the automated workflow can deliver accurate and reproducible predictions even for non-standard and heterogeneous setups, confirming that robustness and accuracy can be maintained simultaneously. The study highlights the potential of CFD4DC to support wider adoption of CFD in data center research and practice and to serve as a foundation for future scientific investigations. • An automation framework for CFD analyses of data centers, CFD4DC, is introduced. • A validation for a static scenario is realized for a two-rack micro data center. • The mean absolute error of 0.63 °C is achieved for k- ɛ turbulence model. • CFD4DC robustness is validated for non-uniform and heterogeneous conditions.