By Arjan Kole
The normal practice of modeling the performance of an air mover in a system is the following:
- The CFD model is built of the application where the air mover is implemented, including the air mover itself.
- The measured performance from testing of the air mover is applied to the model.
- The airflow and/or thermal simulation is run and the operating point of the air mover in the system is then determined.
Typically, the performance of the air mover is characterized when it is attached to a flow bench. The airflow provided by the air mover is measured against the applied back pressure.
|A typical set up is shown in Figure 1 and the measured performance curve is given in Figure 2. In order to verify the accuracy of the model, the standalone model is simulated sitting in ‘free air’, without any obstruction.
|It is observed that the operating point of the air mover is not equal to the free delivery of the air mover. This raises the question on the accuracy of the model. This raises some fundamental ‘issues’ with the approach of modeling prime movers by this approach.
In the experiment, the performance of the air mover is measured, which includes losses at the air intake side, in the air mover itself and at the exhaust side. When the simulation is performed, the losses at both the intake and the exhaust are taken into account again. As a result, the simulation will predict a lower airflow than the actual. Typical airflow pattern is shown in Figure 3.
|The question that rises from this is “what is the implication when this ‘incorrect’ performance curve is used to predict the airflow through a system?” This was investigated by determining the performance curve, corrected for the intake and exhaust side losses. The correction is done by determining the operating pressure at reduced air mover speeds, this gives the operating or system curve of the stand-alone air mover, see Figure 4.
The performance curve can then be corrected for these losses (additional pressure loss is added to the measured performance curve according to the pressure head for each airflow following the system curve).
Figure 5 shows that the correction for the fan curve when it is operating closer to the free delivery airflow is the highest. When the operating point moves to the left, the effect of the correction is lower.
Figure 5: Effect of corrected performance curve on prediction of operating point in the system
This results in the following conclusions:
- Using the measured air mover performance curve is conservative; the corrected performance curve is better.
- High airflow/air speed devices are more prone to have this ‘issue’ than lower airflow/air speed devices as the contribution of the losses at the intake and exhaust due to compression and expansion of the air is small compared to the losses in the air mover itself.
- Depending on the operating point on the performance curve, the impact of the correction maybe significant.