Hamza Salih Erden

Design of Simulated Server Racks for Data Center Research

Conducting experiments on real high-density computer servers can be an expensive and risky task due to the risks associated with unintended inlet temperatures that exceed the server’s red-line temperature limit. Presented herein is the development of the simulated chassis that mimic real computer servers. Briefly, twelve high-power simulated chassis were designed and built to accurately simulate the actual operating conditions of a real computer chassis in a data center. Each simulated chassis is designed to have approximately 300 Pa pressure drop at a flow rate of 600 cfm to represent a real IBM server chassis. Additionally, the simulated chassis are designed to match the thermal mass of a real server. Eight of the simulated chassis were designed to have constant speed fans and variable heating power while the remaining four chassis were designed to have variable speed fans and variable heating power. Further discussions about the design phase of the simulated chassis are the substantial part of this paper. Underlining the challenges and safety issues with high-power chassis, guidelines for designing and constructing a chassis that simulates the real environment of a typical data center are presented.Copyright

Transient Thermal Response of Servers Through Air Temperature Measurements

Transient CFD analysis of data centers requires appropriate representations of the transient thermal characteristics of servers. Thermal conductance and thermal capacitance are two determining characteristics for the response of servers under unsteady conditions. Previous studies proposed tests that require detailed temperature and thermal capacitance measurements for each of the server component, requiring access to individual components inside the server. In this paper, we propose a method for obtaining the transient thermal characteristics of a server from server inlet and outlet temperatures under transient operating conditions.Copyright

Experimental investigation of CRAH bypass for enclosed aisle data centers

Aisle containments in data centers help provide uniform server inlet air temperatures. This allows the cooling system to run at a higher evaporator temperature and more efficiently. On the other hand, CRAH units run at higher speeds to ascertain that racks receive sufficient air flow. Since CRAH fan power already constitutes an important component of data center power use, such increases in the fan power can overshadow the energy savings due to more efficient chiller operation. CRAH bypass configuration is proposed to achieve optimum operating condition for enclosed aisle data centers. This configuration utilizes fan-assisted perforated floor tiles to induce a fraction of tile flow from the room through bypass ports or leakage paths and help decreasing the amount of air flow passing through the large flow resistances of CRAH units. Experimental results show that there is an optimum operating condition for the specific data center test cell that is designed to represent an enclosed aisle data center utilizing the proposed CRAH bypass configuration. Here, the flow characteristics of major system components and experimental measurements have been used to calibrate a flow network model (FNM) for the design optimization and trade-off analysis of the proposed system. Calibrated FNM along with a thermodynamic model (TM) of the cooling infrastructure provides an estimate of the energy use at various fractions of CRAH bypass air and chilled water temperatures. This study introduces the design of the experimental setup for testing CRAH bypass configuration for enclosed aisles and for calibrating models to predict the cooling infrastructure energy saving potential of the proposed technique.

A hybrid lumped capacitance-CFD model for the simulation of data center transients

Transient thermal events in air-cooled data centers may lead to undesirable operating conditions such as the formation of hot spots and associated degradation of equipment reliability. These transients may be caused by cooling equipment failures, server load changes, or other time-dependent scenarios in data center operations. This paper introduces a fast-executing hybrid computational fluid dynamics (CFD)/Lumped-Capacitance model for predicting server inlet temperatures resulting from common transient events such as server shutdown, partial or total chilled water interruption, or partial or total failure of the computer room air handlers (CRAH). The model uses initial steady-state CFD or experimental data in combination with several lumped-capacitance models of the various thermal masses in the data center, including the servers, the room enclosure, the CRAHs and the underfloor plenum. The inclusion of these thermal capacitances and their associated thermal conductance was found to be an important contributor to the overall transient response of the data center air-space. The model predictions have been compared with experimental data obtained in a three-rack data-center test cell and found to agree well with the experimental measurements. Examples of the application of the model to more realistic data center configurations are also given.

Energy and environmental assessment of on-site power and cooling for data centers

High density data centers consume mega watts of electricity for both computing and cooling. Because of the fast paced growth of data center energy demand, effective energy conservation measures must be implemented. The use of on-site combined power generation and cooling systems for data centers that are located on campuses is discussed and compared to grid-powered data centers. The placement of the power plant on site not only reduces transmission losses, but also allows the utilization of the power plant waste heat to generate cooling for the data center and both heating and cooling for adjacent campus buildings. This is why on-site power generation for data centers is an appealing option. Further, with on-site power generation, it is possible to supply the data center with direct current (DC)power, thus, avoiding the cascade of waste in the multiple AC/DC conversions typical of conventional data centers. This article presents a detailed modeling effort to assess the efficiency and environmental and energy cost benefits of such on-site co-generation systems for campus-embedded data centers by utilizing a primary energy analysis.

Energy assessment of CRAH bypass for enclosed aisle data centers

Temperature non-uniformities in traditional data centers can be eliminated or at least reduced by utilizing containment systems. As all servers receive the same inlet air temperature in a contained configuration, the cooling system can be operated more efficiently at a higher temperature, which also increases the potential for free cooling through various economizer modes. However, enclosed aisle configurations require computer room air handler (CRAH) fans to operate at a higher speed and provide entire rack air flow through the perforated tiles, unlike open aisle data centers that can make up a fraction of server air from the data center air space. Hence, the traditional enclosed aisle configuration is likely to consume more fan power. This study confirms that enclosing the aisle does not guarantee optimum cooling infrastructure power in air cooled data centers. Proposed CRAH bypass configuration for enclosed aisle data centers provides a fraction of the tile airflow rate through a set of bypass fans while CRAH fans operate at lower speeds. These low-lift fans operate across a pressure difference between the room and plenum, which is significantly less than the flow resistance of CRAH units. Meanwhile, CRAH fans operate at lower speeds and consume less energy. Accordingly, a certain bypass air fraction with respect to total rack air flow rate leads to a minimum cooling infrastructure power for a particular configuration. This study investigates energy savings potential of the enclosed aisle data centers with CRAH bypass configuration utilizing a calibrated flow network model for estimating the energy consumption of air movers as well as a thermodynamic modeling tool to evaluate the off-design performance of major components of data center cooling infrastructure. Hour-by-hour annual energy simulations complement the energy assessment for 7 U.S. cities considering indirect air side economizer operation.

Room-Level Transient CFD Modeling of Rack Shutdown

Servers have time constants on the order of several minutes, which is significantly longer than many other time constants observed in data centers. When server thermal masses are not negligible, transient scenarios cannot be modeled appropriately with the steady state boundary conditions. By using the time constant and thermal masses of the servers, transient server air exit temperature can be calculated by solving unsteady heat balance equations for the server and air stream. This paper presents a room-level CFD case study of rack shutdown in which transient server exit air temperature are introduced via user defined functions. The results include verification of simulation results with the experimental data taken in a 3-rack test cell.Copyright

Experimental Demonstration and Flow Network Model Verification of Induced CRAH Bypass for Cooling Optimization of Enclosed-Aisle Data Centers

Aisle containment separates hot and cold aisles in air-cooled data centers to prevent hot air recirculation into cold aisle and decrease temperature nonuniformity among the servers. Uniform server inlet temperatures allow the cooling system to operate more efficiently at a higher evaporator temperature. Lower cooling power consumption can be achieved by optimizing the combined power consumption of the chillers and the computer room air handlers (CRAHs) fans. CRAH bypass (BP) is proposed, in which additional fans at the tiles induce a fraction of tile flow from the room into the plenum through low-resistance ports or leakage paths, thus decreasing the airflow passing through the high pressure resistance of the CRAH heat exchangers and filters and associated fan power usage. A further advantage of the proposed induced CRAH BP is the elimination of leakage as a result of reversing the room-plenum pressure difference. However, in order to keep the enclosed-aisle temperature below acceptable limits (typically ≤27 °C), the chiller needs to operate at a lower, less efficient temperature. This paper presents the experimental verification of a flow network model (FNM) in a data center test cell demonstrating the proposed induced CRAH BP concept. Exercise of the verified FNM in conjunction with a thermodynamic model of the cooling infrastructure (TDM) confirms that there is an optimum BP fraction that minimizes the combined chiller and CRAH fan power consumption.

Parameter Estimation for Lumped Capacitance Modeling of CRAH Units During Chilled Water Interruption

Lumped capacitance models have been introduced to study transient thermal response of data centers. Chilled water interruption of a Computer Room Air Handling (CRAH) unit is one of several failure scenarios of data center cooling infrastructures. In such a scenario, predicting the transient thermal response of the CRAH unit depends requires the determination of the CRAH lumped capacitance model parameters: the thermal capacitance (thermal mass) and the time constant. In this paper, we propose an experimental methodology to extract sufficient information for the lumped capacitance modeling of CRAH units. The method requires measurements of inlet and exit air temperature, air flow rate and CRAH fan power. If the chilled water supply to a CRAH unit is intentionally interrupted in a data center with multiple redundant CRAH units, sufficient information to estimate the CRAH lumped capacitance parameters can be obtained without disturbing the data center operation.Copyright

Development of an IT equipment lumped capacitance parameter database for transient data center simulations

The transient behavior of IT equipment can be represented by a lumped thermal capacitance for use in data center level transient thermal simulations. Previous work has proposed a mathematical methodology to extract the necessary lumped capacitance parameters, the servers time constant and heat transfer effectiveness, via air temperature measurements. This work describes and experimentally tests that proposed methodology to introduce a possible approach for developing an easy-to-use database of lumped capacitance characteristics for use in transient thermal simulations. The database will allow the user to select the appropriate transient parameters based on characteristics that do not require an “autopsy” of each-and-every server in the data center. Experiments are conducted to provide representative transient parameters which classify the servers by mass density and operating air flow rate. The paper describes the experimental methodology in detail to allow for the easy addition of other IT equipment or future server generations.