Top > Application Examples > National Defense Academy of Japan

### National Defense Academy of JapanCombining Tests and CFD Analysis to Solve Long Standing Issues

The configuration of cooling fans inside electronic devices is complex. This frequently results in generating large amounts of data when trying to represent the configuration accurately for fluid analysis calculations. To deal with this, engineers use the concept of the pressure-flow-rate (PQ) characteristics to simplify the calculation, although it has been pointed out that this simplification often becomes the source of errors. Professor Hajime Nakamura, from the National Defense Academy of Japan, has been exploring these issues pertaining to fan models for electronic devices and has discovered a more accurate model.

Professor Hajime Nakamura
National Defense Academy of Japan
Mechanical Engineering Dept.

#### Verifying the Best Definition of Rotational Component

The next concern was the unpredictable velocity distribution. Ideally the air from the fan would be defined with two components, axial velocity and tangential velocity. Unfortunately, only the flow rate is known from the PQ curve and not the rotational component. Until recently engineers using CFD applied different coeffcients to account for the rotational velocity component. Swirl ratio, defined by the rotational and axial velocity, was commonly used. The ratio of angular velocity and impeller rotational velocity was also used in older versions of Cradle’s software. Determining the proper value for the rotational component is a complicated task; it requires significant engineering experience and insight. Whether such approaches are truly appropriate for accurately representing the actual phenomena is still not fully verified.

Figure 3: Comparison between experimental test and
analysis results of rotational fan model (from left: analysis results of fan model, smoke-wire visualization,
​comparison of velocity distributions). Click to enlarge.

Professor Nakamura started to consider applying a model, which calculates rotational components using the number of vanes, number of rotations, inner radius, outer radius, and thickness of a vane. He based his theory on the concept of drag coefficient (CD). The drag coefficient varies depending on the model shape and follows the law of similarity. For example, the drag coefficient of any disk normal to the flow is roughly 1.2, which remains constant unless the disk is extremely small. Professor Nakamura thought that this idea could be applied to rotating vanes, regardless of their size and velocity. He calculated the rotational force coefficient as CΘ using the rotational force expressed as FΘ=1/2ρuΘ2ACΘ. He found that a constant CΘ of 0.6 gave analysis results that best matched experimental test results (shown in Figure 3), and worked for both large and small fans. Professor Nakamura also found that for the smallest fan of □17mm tested, the optimum value of CΘ was somewhat higher although the differences of the results were still relatively small compared to using CΘ=0.6.

Figure 4: Changes by radial positions of rotational force coefficient, swirl ratio, and ratio of angular velocity (validated by the detailed analyses where influence of impeller configuration and its rotation was taken into consideration). Click to enlarge.

Determining the rotational component of the velocity had always been a challenge for Professor Nakamura as he had not yet been able to identify the best model for this calculation. To overcome the challenge, he started using rotational force coefficients as an alternative to applying the drag coefficient. He was concerned whether the idea of force coefficient should be applied to rotating vanes because drag coefficients are normally used for modeling a cylinder or sphere in a uniform flow. Yet the results when using Cθ were convincing. Professor Nakamura has also evaluated swirl ratio and the ratio of angular velocity to rotational velocity, and found that using Cθ works better as a constant value from the center to outside of the vane (Figure 4). Based on the investigations, he named Cθ the non-dimensional swirl coefficient, which is now used in the Software Cradle fan model. As a result, engineers can directly reap the benefits of Professor Nakamura’s fan research when they use Cradle software.

#### Analyzing a Highly Packed Device

Professor Nakamura’s last concern was that the change in the PQ curve becomes significant when the device was more densely-packed. To find a solution he examined the application of the revised fan model. When blockages got close to a fan, the PQ curve changed as the blockages restricted the fluid flow. This was caused by the additional pressure loss solely due to the presence of obstacles. Namely, the fan performance itself did not change by the obstruction. The only time that analysis and test results did not match was when obstacles were too close to the upstream side of the fan. In this case, the flow's entrance angle altered, which deteriorated the fan performance. As long as this condition is not reached, Professor Nakamura’s model can be used to represent fluid flow.

#### Further Investigation Using Axial Fans

“I hope to investigate different kinds of fan going forward,” says Professor Nakamura. This includes a fan with one rotor with a non-rotating deflector just downstream of the rotor, and a type with two counter-rotating rotors in a series. He suspects that the revised fan model can be used to analyze a fan with the deflectors. “I think the major issues with fan models have mostly been solved. I’m expecting that manufacturers will continue to develop more capable fan designs, but my basic fan model will be sufficient, as long as we review the coefficient each time,” he says.

Professor Nakamura’s next challenge beyond optimizing fan models is the modeling of heat transfer. His research interest includes investigating heat transfer with fluid flows, which he believes can be analyzed accurately using CFD software when the flow is laminar. It becomes more difficult to accurately simulate when the flow becomes turbulent. In this case, the flow and temperature fluctuate in complicated manner, which directly impact the temperature fluctuation in the solid. Professor Nakamura hopes his work can lead to improved heat transfer modeling between the fluid and solid for turbulent flows.

#### CFD Analysis as a Necessary Tool to Verify Tests

Just as Professor Nakamura’s research has always done, he will continue to explore and verify phenomena through experimental tests that challenges the capabilities of traditional CFD analysis. This does not mean that Professor Nakamura is against CFD analysis; in fact, he considers CFD a necessary tool to verify tests. He recommends combining CFD analysis with actual tests. This is what he did with the fan model, and adopting this approach will help convince other engineers to experiment. Regarding Cradle’s software, Professor Nakamura says “I have great faith in their software. I’m also reassured by how quickly they implemented the research results into their software.” In particular, many engineers found that the software’s capability to adjust the dynamic pressure was useful. “Their feedback helps our research continue to make progress. This, in turn, can lead to new investigations, which is the pathway to new possibilities,” says Professor Nakamura.
* All product and service names mentioned are registered trademarks or trademarks of their respective companies.
* Contents and specifications of products are as of June 1, 2013 and subject to change without notice. We shall not be held liable for any errors in figures and pictures, or any typographical errors.

#### Biography

Professor Hajime Nakamura,
National Defense Academy of Japan
Mechanical Engineering Dept.
Thermal Engineering Laboratory
Research interests Thermal Engineering, Fluid Engineering Graduated Mechanical Science and Engineering from Tokyo Institute of Technology, and attained M.S. and Ph.D. degrees in Energy Science from Tokyo Institute of Technology Ph.D. in Engineering

#### Download

This article is also available in pdf

#### Featured Software

General Purpose Structured Mesh Thermal-fluid Analysis Software
More Details

3D Structured Mesh Thermal-fluid Analysis System for Electronics Cooling
More Details

General Purpose Unstructured Mesh Thermal-fluid Analysis System
More Details

### Visitors also read

Azbil Corporation

Achieving Size Miniaturization and Noise Reduction of Control Valves

Murata Machinery, Ltd.

Fluid Analysis Software Helps Demonstrate Product Mechanism to Customers

Tokai University

Students Take the Lead in Simulating Solar Car Aerodynamics

SAXA Inc.

Key to Success: Creating an environment in which engineers can objectively evaluate analysis results

#### Inquiry

Contact us from the inquiry form below for any inquiry regarding this article.