Passive cooling solutions for HPC data centres
This final-year project investigates the implementation of passive cooling techniques in high-performance computing (HPC) data centres, using Computational Fluid Dynamics (CFD) to analyse airflow dynamics and thermal management. The study focuses on designing server rack configurations with and with...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Final Year Project |
Language: | English |
Published: |
Nanyang Technological University
2024
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/177488 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | This final-year project investigates the implementation of passive cooling techniques in high-performance computing (HPC) data centres, using Computational Fluid Dynamics (CFD) to analyse airflow dynamics and thermal management. The study focuses on designing server rack configurations with and without blanking panels and cable gland seals, examining their impact on cooling efficiency and energy consumption.
Through simulations modelled using SolidWorks and conducted using the student version of ANSYS Fluent 2024 R1, various design modifications are tested to ascertain their effectiveness in enhancing passive cooling strategies despite limitations in mesh resolution.
The findings indicate that strategic placement of blanking panels significantly reduces temperature elevations and prevents the formation of hotspots. Using cable gland seals minimises unintended air leakages, enhancing the stability and control of the cooling environment. The research confirms the potential of passive cooling systems to improve the reliability and longevity of server equipment by maintaining ideal operating temperatures and minimising thermal stress.
This project contributes to understanding sustainable cooling practices in data centres, aligning with global efforts to reduce energy use and carbon footprints. Recommendations for future work include further refining the CFD models using professional-grade ANSYS software, such as an Educator or Researcher license, and expanding the research to encompass more varied data centre environments, ensuring the scalability and practical applicability of passive cooling solutions. |
---|