CFD - Based Investigation of Effects of Obstruction in Front of Small Axial Cooling Fan and Deterioration of Supply Flow Rate

This study describes the deterioration of a small axial fan’s supply flow rate in high-density packaging electronic equipment. A cooling fan flow rate can be predicted by its P-Q curve, which shows a relationship between a pressure rise at a fan (ΔP) and a supply flow rate (Q). However, in high-density packaging electronic equipment, the fan performance is affected by the mounting components around the fans, and the accurate prediction of the supply flow rate becomes difficult. This paper tried to do flow visualization around a small axial cooling fan’s impellers when the obstruction was mounted in front of the fan through CFD analysis. A relationship between the supply flow rate by the fan and the flow pattern around the impellers was investigated while changing the distance between the test fan and the obstruction. Through this study, the following results can be obtained. The fan’s flow is stable in the rotating stall region and the higher flow rate operating points regardless of whether or without the obstruction. At the lower flow rate conditions, the formation of a complex unsteady flow is reproduced. As the flow rate decreases, the flow’s separation point becomes closer to the leading edge of the impeller. In the case of obstruction, the change of the flow pattern causes a larger attack angle. As a result, fan performance is degraded.

Noise Reduction Analysis of Electronic Device Cooling Fan with Duct and Its Application Under Variable Working Conditions

A noise reduction method for axial flow fans using a short inlet duct is proposed. The pattern of noise reduction imposed by the short inlet duct on the axial flow cooling fan under variable working conditions was experimentally and numerically examined. A 2-cm inlet duct was found to reduce tonal noise. As the tip Mach number of the fan increased from 0.049 to 0.156, the reduction in the total average sound pressure level at 1 m from the fan increased from 0.8 dB(A) to 4.3 dB(A), and further achieved 4.8 dB(A) when a 1-cm inlet duct was used. The steady CFD showed that the inlet duct has little effect on the aerodynamic performance of the fan. The results of the unsteady calculation showed that the suction vortexes move upstream to weaken the interaction with the rotor blades, which significantly reduces the pulsating pressure on the blades. The SPL at the BPF contributed by the thrust force was calculated to reduce by 36 dB at a 135° observer angle, reflecting the rectification effect of the duct on the non-uniform inlet flow. The POD of the static pressure field on the blades verified that the main spatial mode is more uniformly distributed due to the duct, and energy owing to the rotor-inlet interaction decreases. A speed regulation strategy for the cooling fan with short inlet duct is proposed, which provides guidance for the application of this noise reduction method.

Pressure Drop Mechanisms Generated in a Cooling System Enclosure of Construction Machinery

A potential way to reduce cooling system noises generated by heavy construction machines is to generate the required cooling airflow with a low fan speed, and one way to accomplish this is to optimize the ventilation path through which the airflow generated by the cooling fan must travel. However, while the computational fluid dynamics (CFD) approach would be effective for modeling the three-dimensional (3D) pressure drop characteristic of such systems, there have been few reports aimed at clarifying the loss generation mechanisms or suggesting minimization methods based on flow field viewpoints. Accordingly, in this study, we visualize the 3D flow field characteristics of an electric cooling fan system installed within the cooling enclosure of a heavy construction machine and investigate the details of the system’s pressure drop mechanisms. Our results confirm that airflow pressure declines in areas other than the radiator account for more than half of the reduced pressure experienced by the whole system. Additionally, we found that, in the exhaust side enclosure, pressure drops increased because the exhaust port outlet shapes were not optimized to the annular airflow of the cooling fan. Most notably, we found that in the region before reaching the exhaust port outlets, the airflow from the fan repeatedly collides with obstacles within the enclosure, thus producing stagnation and turbulence that exacerbates pressure drops before being expelled into the outside environment.

Numerical simulations of flow induced noise from a dual rotor cooling fan used in electronic cooling systems

Hard Disk Drive (HDD) system enclosures in a data center require effective cooling systems to avoid HDD overheating. These systems often rely on air cooling because of their cost effciency and maintainability. Air cooling systems typically consist of an array of axial fans which push or pull the air through the system. These fans emit high level tonal noise particularly at high tip speed ratios. High-capacity HDDs, on the other hand, are sensitive to high acoustic noise, which consequently increases the risk of read/write error and deteriorates drive performance. Therefore, cooling fan noise adversely affects the function of the HDD enclosure systems which emphasizes the need to understand the noise sources and develop methods to mitigate HDD noise exposure.

Two-step computational aeroacoustics approach for underhood cooling fan application

Aeroacoustic noise is one of the important characteristics of the fan design. Computational Aeroacoustics (CAA) can provide better design options without relying on physical prototypes and reduce the development time and cost. There are two ways of performing CAA analysis; one-step and two-step approach. In one-step CAA, air flow and acoustic analysis are carried out in a single software. In two-step approach, air flow and acoustic analysis are carried out in separate software. Two-step CAA approach can expedite the calculation process and can be implemented in larger and complex domain problems. For the work presented in this paper, a mockup of an underhood cooling fan was designed. The sound pressure levels were measured for different installation configurations. The sound pressure level for one of the configurations was calculated with two-step approach and compared with test data. The compressible fluid flow field was first computed in a commercially available computational fluid dynamics software. This flow field was imported in a separate software where fan noise sources were computed and further used to predict the sound pressure levels at various microphone locations. The results show an excellent correlation between test and simulation for both tonal and broadband components of the fan noise.