Experimental Control of Thermal Processes Inside the Datacenter

Optimizing the thermal processes inside data centers in order to reduce the energy consumption and lowering the environmental footprint was always a big challenge.Finding solutions to improve the current cooling system and reduce the energy inefficiency can be achieved with various approaches.The purpose of this work was to improve the thermal processes in the INCDTIM’s data center by developing a smart cooling system that provides localized cooling when and where is needed. For this, experimental methods were used, implementing both, passive solutions (pipes, optimizations of the plenum) and active solutions (active grills for more efficient air control, remote control automation). First, we started from a 3D model of the data center, followed by the simulations of the actual thermal conditions in Ansys CFD (Computational Fluid Dynamics) software module. These simulations identified the problems, and the cold pockets and provided a tool to deliver a finite product that will significantly improve the cooling system. After this, a comparison was made with the improved thermal conditions in order to prove that the physical model matches the theoretical one. After that, the automation needs to be built in order to control the thermal processes. The experimental connection of the PLC to a cloud-based platform aims to monitor and control the thermal processes in the Cloud.