Fan trays are assemblies of 2, 4 or 6 compact fans mounted on a flat tray and normally installed within electronic enclosures. They are particularly good at assisting airflow through an electronic enclosure and providing spot or large area cooling. They are also used at the top of cabinets, within the lid of the enclosure, to extract air out of the cabinet. But does the fan tray cooling solution really work in this type of application?
To find out, systems providers at ebm-Ziehl UK decided to conduct a series of tests using a variety of fan configurations
The Papst 4550N is typical of the type of fan used in a fan tray. It produces 145m3/hr at free air. Fan law theory states that where two or more fans receive air from – and deliver air into – a common system, they operate in parallel.
Following the theory, the volume flow is multiplied by the number of fans to give the total volume. In this example, two 4550N compact fans in parallel would provide 2 x 145m3/hr = 290m3/hr at free air.
This theory would only hold true if there were a large separation between the fans. In reality, however, the fans are located adjacent to each other and fight to suck in the air that is located between the two fans. This restricts the performance that the fans can achieve.
In the case of two compact fans side by side, the actual volume flow is in the region of 10% less than the theoretical volume flow.
Engineers at ebm-ZIEHL constructed an airflow test rig to BS848: Part 1 to determine the true performance of the six-fan tray application and also to analyse its operation in the roof of a 19 inch enclosure. The purpose of the evaluation was to look at alternative centrifugal fans.
In this case, as stated earlier, all six fans are fighting to suck in the same molecules of air.
Theory states that the free air volume should be 6 x 145m3/hr = 870m3/hr. The reality is significantly less. Through investigation on the airflow test rig, it was found that this layout can only achieve 390m3/hr, which is less than half of the theoretical volume flow.
After realising the shortcomings of the fan tray following the test procedure, the tray was then mounted into the roof of the cabinet.
The close proximity of the roof to the compact fans causes high turbulence and therefore high resistance to flow. It was found that, in this layout, the volume flow was further reduced from 390m3/hr down to 160m3/hr with the roof fitted to the cabinet. And this is before any extra resistance to flow is added within the electronic enclosure. Volume flow dropped from a theoretical 870m3/hr down to an actual 160m3/hr.
There has been debate on whether one single backward curved centrifugal fan could be used instead of six compact fans. To find out, a fan tray using a 180mm diameter backward curved impeller was constructed and the performance compared to the six-compact fan tray when installed in the roof of the cabinet.
Three measurements using the backward curved fan were taken. The first was at full voltage to determine the maximum volume flow in the application. The second was taken with a reduced voltage to find a volume flow of the backward curved fan at an equivalent noise level of the six compact fans. Finally, a third measurement was taken where the backward curved fan was set at a volume flow equivalent to that of six compact fans.
One backward curved fan?
It was found that the backward curved fan produced 94% more volume of air flow than the six compact fans for the same noise level. The backward curved fan produced 310m3/hr compared to the six compact fans of the combined output of 160m3/hr.
The investigation showed that the single backward curved fan is also a quieter solution than the six compact fan option, providing almost twice the capacity at the same noise level.
More importantly, it can provide extra performance to overcome the increased losses that will occur when the electronic enclosure is populated.
The six compact fan tray would be unable to keep maintaining the required volume flow, as it is working at full speed and has no further capacity.