Fan performance rating tests are conducted in standard configurations which, unfortunately, are not always exactly reproduced in field installations. The connection between a fan inlet and the air system can have a disruptive effect on the airflow into the fan inlet. Non-uniform airflow into the inlet is the most common cause of deficient fan performance.

An elbow or a 90-degree duct turn located at the fan inlet will not allow the air to enter uniformly and will result in turbulent and uneven airflow distribution to the fan impeller. Air has mass, and a moving airstream has momentum. Therefore, the airstream resists changing direction within an elbow (Figure 1).

The effect of these airflow disturbances, called "system effects" by Air Movement and Control Association International, Inc. (AMCA), is nowhere more disruptive than in cases where a vortex or spin is produced at the inlet of a centrifugal fan. Figure 2 shows an example of a spin-generating connection.

Effects Of Fan Inlet Spin

At a fan inlet, air spinning in the same direction as the impeller rotation reduces the fan pressure-volume curve by an amount dependent upon the intensity of the vortex. The effect is similar to the change in the pressure-volume curve achieved by inlet vanes installed at a fan inlet, which induce a controlled spin and so vary the volume airflow rate of the system.

A counter-rotating vortex at the inlet will result in a slight increase in the pressure-volume curve, but power consumption will increase substantially. There are occasions, with counter-rotating swirl, when the loss of performance is accompanied by a surging airflow. In these cases, the surging may be more objectionable than the performance change.

The cause of inlet spin is not always obvious. For example, cyclone separators are designed to take advantage of a spinning airstream, which may extend downstream several duct diameters to the fan inlet. Some common duct connections, which cause inlet spin, are shown in Figures 3 and 4.

Inlet spin is one possible side effect of using rectangular ducts or adjacent elbows (Figure 3). The reduction in capacity and pressure (or system effect) for this type of inlet condition is impossible to tabulate. (System effect factors have been tabulated for single round and for single square elbow ducts in AMCA Publication 201, Fans and Systems.) The many variations in width and depth of the duct influence the reduction in performance to varying degrees and, therefore, this inlet design should be avoided. Capacity losses as high as 45% have been observed.

When capacity losses reach the magnitude of 45%, the application is in serious trouble. The air system problem must be corrected. This can lead to a lot of lost time by all involved. Fingerpointing, which can reach new heights in such circumstances, can be avoided by installing round or square elbow ducts with tabulated system effect factors.

(By the way, Figure 4 shows an eccentric converging branch. The eccentricity would not be noticed without close inspection of the ductwork. Once the eccentric nature of the branch exit is observed, the spinning airflow is obvious.)

Corrections For Inlet Spin

The ideal inlet airflow condition is one that allows the air to enter axially and uniformly without spin in either direction. The ideal connection would include up to 10 diameters of straight duct leading to the fan inlet depending on the air velocity.

Where space limitations prevent the use of optimum fan inlet connections, more uniform airflow can be achieved by the use of turning vanes in the inlet elbow (Figure 5), or splitters in the duct (Figure 6).

Regarding inlet turning vanes, numerous variations of turning vanes are available, from a single, curved, sheet metal vane to multi-bladed "airfoil" vanes. The pressure drop through these devices must be added to the system pressure losses. The elbow manufacturer publishes these losses, but it should be realized that the cataloged pressure loss would be based upon uniform airflow at the entry to the elbow.

If the airflow approaching the elbow is significantly non-uniform because of a disturbance farther upstream in the system, the pressure loss through the elbow will be higher than the published figure. The effectiveness of the vanes in the elbow will also be reduced.

In regard to splitters, Figure 4 shows an eccentric converging branch. The recommended correction shown in Figure 6 takes advantage of a splitter, but equally important, the branch was relocated to the duct centerline, a position normally expected using good duct design practices.

Splitter variations range from a single divider sheet to multiple pie-shaped dividers up to four duct diameters long (Figure 7). In most cases the final design will depend upon airflow and power requirements. If the loss of performance is small, it may be easier to increase the fan speed to compensate.

When airflow and power are critical, a complete redesign of the inlet duct may be required. Most cases can be dealt with by inserting splitters to correct the major problem, though the splitter adds to the system pressure loss. ES