Emad Samadiani

Proper Orthogonal Decomposition for Reduced Order Thermal Modeling of Air Cooled Data Centers

Computational fluid dynamics/heat transfer (CFD/HT) methods are too time consuming and costly to examine the effect of multiple design variables on the system thermal performance, especially for systems with multiple components and interacting physical phenomena. In this paper, a proper orthogonal decomposition (POD) based reduced order thermal modeling approach is presented for complex convective systems. The basic POD technique is used with energy balance equations, and heat flux and/or surface temperature matching to generate a computationally efficient thermal model in terms of the system design variables. The effectiveness of the presented approach is studied through application to an air-cooled data center cell with a floor area of 23.2 m 2 and a total power dissipation of 240 kW, with multiple turbulent convective components and five design variables. The method results in average temperature rise prediction error of 1.24°C (4.9%) for different sets of design variables, while it is ∼150 times faster than CFD/HT simulation. Also, the effects of the number of available algebraic equations and retained POD modes on the accuracy of the obtained thermal field are studied.

Adaptable Robust Design of Multi-Scale Convective Systems Applied to Energy Efficient Data Centers

A design approach is presented to bring adaptability and robustness to multi-scale convective systems. The design method is centered on proper orthogonal decomposition-based reduced order thermal modeling, robust design principles, and the compromise decision support problem construct. The method application for an energy efficient air-cooled data center guarantees safe and robust operation over a 10-year period of increasing power dissipation. The data center traditional design consumes 14–85% more energy than the adaptable design over the 10 years. A robust solution is also found to reduce the variability in the thermal response by 80% compared with an optimal solution.

Reduced order thermal modeling of data centers via proper orthogonal decomposition: a review

Purpose – The purpose of this paper is to review the available reduced order modeling approaches in the literature for predicting the flow and specially temperature fields inside data centers in terms of the involved design parameters.Design/methodology/approach – This paper begins with a motivation for flow/thermal modeling needs for designing an energy‐efficient thermal management system in data centers. Recent studies on air velocity and temperature field simulations in data centers through computational fluid dynamics/heat transfer (CFD/HT) are reviewed. Meta‐modeling and reduced order modeling are tools to generate accurate and rapid surrogate models for a complex system. These tools, with a focus on low‐dimensional models of turbulent flows are reviewed. Reduced order modeling techniques based on turbulent coherent structures identification, in particular the proper orthogonal decomposition (POD) are explained and reviewed in more details. Then, the available approaches for rapid thermal modeling of...

Numerical Modeling of Perforated Tile Flow Distribution in a Raised-Floor Data Center

This paper is centered on quantifying the effect of computer room and computer room air conditioning (CRAC) unit modeling on the perforated tile flow distribution in a representative raised-floor data center. Also, this study quantifies the effect of plenum pipes and perforated tile porosity on the operating points of the CRAC blowers, total CRAC air flow rate, and its distribution. It is concluded that modeling the computer room, the CRAC units, and/or the plenum pipes could make an average change of up to 17% in the tile flow rates with a maximum of up to 135% for the facility with 56% open tiles while the average and maximum changes for the facility with 25% open tiles are 6% and 60%, respectively.

Reduced Order Thermal Modeling of Data Centers via Distributed Sensor Data

In this paper, an effective and computationally efficient proper orthogonal decomposition (POD) based reduced order modeling approach is presented, which utilizes selected sets of observed thermal sensor data inside the data centers to help predict the data center temperature field as a function of the air flow rates of computer room air conditioning (CRAC) units. The approach is demonstrated through application to an operational data center of 102.2 m2 (1100 square feet) with a hot and cold aisle arrangement of racks cooled by one CRAC unit. While the thermal data throughout the facility can be collected in about 30 min using a 3D temperature mapping tool, the POD method is able to generate temperature field throughout the data center in less than 2 s on a high end desktop personal computer (PC). Comparing the obtained POD temperature fields with the experimentally measured data for two different values of CRAC flow rates shows that the method can predict the temperature field with the average error of 0.68 °C or 3.2%. The maximum local error is around 8 °C, but the total number of points where the local error is larger than 1 °C, is only ∼6% of the total domain points.

Coordinated Optimization of Cooling and IT Power in Data Centers

Concurrency and exchanging design knowledge among thermal and IT management are required to achieve an energy efficient operational data center. In this paper, a design approach is presented to bring adaptability and concurrency for coordinated minimization of cooling and IT power consumption in data centers. The presented approach is centered on a proper orthogonal decomposition based reduced order thermal modeling approach, and power profiling of the IT equipment to identify the optimal parameters of the air cooling systems along with optimal dynamic workload distribution among the servers. The method is applied to a data center cell with different rack and server architectures. The results show that the design approach results in 12-70% saving in the total energy consumption of the data center cell for various scenarios, compared with a baseline design.

The Thermal Design of a Next Generation Data Center: A Conceptual Exposition

In the near future, electronic cabinets of data centers will house high performance chips with heat fluxes approaching 100 W/cm and associated high volumetric heat generation rates. With the power trends in the electronic cabinets indicating 60 kW cabinets in the near future, the current cooling systems of data centers will be insufficient and new solutions will need to be explored. Accordingly, the key issue that merits investigation is identifying and satisfying the needed specifications of the new thermal solutions, considering the design environment of the next generation data centers. Anchoring our work in the open engineering systems paradigm, the authors identify the requirements of the future thermal solutions and explore various design specifications of an ideally open thermal solution for a next generation data center. To approach an open cooling system for the future data centers, the concept of a thermal solution centered on the multi-scale (multilevel) nature of the data centers is discussed. The potential of this solution to be open, along with its theoretical advantages compared with the typical air cooling solutions is demonstrated through some scenarios. The realization problems and the future research needs are highlighted to achieve a practical open multi-scale thermal solution in data centers. Such solution is believed to be the most effective and efficient for the next generation data centers

Exploring the Advantages of Materials Design in a Product Design Process

In this paper, we explore the benefits of materials design in a product design process. We also compare the methods of material selection and materials design by demonstrating two examples—the design of a cantilever beam for minimum weight and the design of a fan blade for minimum weight. The design of the cantilever beam is carried out using Ashby’s material selection method as well as a proposed method for materials design. The design of the fan blade and its material is completed using computational tools. Our goal in this paper is to demonstrate the benefits of materials design over material selection methods and to illustrate the flexibility inherent in materials design processes. We are more interested in revealing the possibilities of materials design, rather than the specific results from the example problems. The investigation of materials design presented in this paper moves us one step closer towards the realization of a systematic, inductive method for the concurrent design of products and materials.Copyright

Reduced Order Modeling Based Energy Efficient and Adaptable Design

In this chapter, the sustainable and reliable operations of the electronic equipment in data centers are shown to be possible through a reduced order modeling based design. First, the literature on simulation-based design of data centers using computational fluid dynamics/heat transfer (CFD/HT) and low-dimensional modeling are reviewed. Then, two recent proper orthogonal decomposition (POD) based reduced order thermal modeling methods are explained to simulate multiparameter-dependent temperature field in multiscale thermal/fluid systems such as data centers. The methods result in average error norm of ~6% for different sets of design parameters, while they can be up to ~250 times faster than CFD/HT simulation in an iterative optimization technique for a sample data center cell. The POD-based modeling approach is applied along with multiobjective design principles to systematically achieve an energy efficient, adaptable, and robust thermal management system for data centers. The framework allows for intelligent dynamic changes in the rack heat loads, required cooling airflow rates, and supply air temperature based on the actual momentary center heat loads, rather than planned occupancy, to extend the limits of air cooling and/or increase energy efficiency. This optimization has shown energy consumption reduction by 12–46% in a data center cell.