With the rapid growth of AI and Big Data, reducing energy consumption in data centers has become a significant challenge. Small-scale server rooms, which often depend on standalone cooling units instead of specialized infrastructure, are especially vulnerable to airflow inefficiencies and localized hotspots. While traditional HVAC theory can estimate total cooling loads, it does not effectively predict the local thermal distributions that depend on equipment placement. This study employs Computational Fluid Dynamics (CFD) to address these limitations. It focuses on a server room containing five HPCs and three cooling units, comparing theoretical HVAC calculations with 3D thermal fluid analysis conducted using Ansys Fluent. The research evaluates the thermal performance of the existing layout (Model 1) and suggests an improved configuration (Model 2). The findings reveal that strategic equipment placement can eliminate hotspots and stabilize operations while reducing the necessary cooling capacity. This research provides a practical framework for enhancing energy efficiency in small-scale server environments.
A Transformer model to predict the remaining useful life of a fuel cell, which has demonstrated superior performance in analyzing time series data. The dataset was created from long-term performance evaluation experiments conducted in rated power mode, with measurements taken every 10 hours. We preprocessed the raw data using a moving average, allocating 70% for training and 30% for evaluation. The model's performance, evaluated through MAE, MSE, and MAPE, was excellent. The fuel cell's critical voltage, defined as 94.5% of its initial voltage, was measured at 0.719 V. During the experimental run, the actual critical time was 106.6 hours, while the model predicted 106.8 hours, resulting in a 0.19% error. Since the predictions were based on data collected up to 93 hours, the estimated remaining life was 13.8 hours.
First
Prev
