Characterization of the Flow and Pressure Drop in Honeycomb Inlets With Grazing Flow

Experiments were conducted in a low speed laboratory wind tunnel in order to characterize the flow and pressure drop through a honeycomb filter inlet when the flow approaches at 90° to the inlet normal, a so-called “grazing flow”. The honeycomb was baselined in a traditional orientation normal to the flow direction. The overall total and static pressure drops and the downstream velocity field were measured using pressure probes and wall static taps upstream and downstream of the honeycomb. The wind tunnel was then re-configured so that the main wind tunnel flow resembled an infinite free-stream approaching the honeycomb inlet at 90° from the normal. Total and static pressure surveys were performed in the airflow upstream of the honeycomb, and at the honeycomb inlet and exit. Single hot wire anemometer surveys of the time mean total velocities were performed at the honeycomb inlet and at various locations downstream of the honeycomb inlet. Because the honeycomb exit flow has a single velocity component, the measured inlet and exit total velocities were sufficient to determine the local approach angle of the inlet velocity vector. It was determined that the radius of curvature of the flow stream entering the honeycomb is small in the upstream region of the honeycomb, and increases monotonically in the flow direction. This largely inviscid effect causes a drop in the static pressure in the upstream region of the honeycomb followed by a gradual recovery of the static pressure in the flow direction. The local mass flow ingested through the honeycomb is low in the forward section and increases monotonically in the flow direction. This non-uniformity in the ingested mass flow results from the non-uniformity in the turning radius and entering static pressure, an inviscid mechanism, and from non-uniformity of the pressure loss coefficient over the honeycomb panel, a viscous mechanism. The total pressure loss coefficient, K, is found to vary from a value of roughly 1.2, for grazing flow approaching at nearly 90° in the forward section of the honeycomb, to 0.9, for a flow approaching from the direction normal to the inlet. It is hypothesized that the increased total pressure loss for the grazing flow is due to flow separation at the inlet of each honeycomb passage.