6.1 Traverse Locations, Disturbances, and Point Layout

Traverse Location and Point Layout

Why a traverse needs many points

Velocity is not uniform across a duct. Wall friction slows the boundary layer, fittings create gradients and swirl, and fan discharge can produce concentrated jets. A single centerline reading therefore cannot represent total flow. A duct traverse samples defined locations across the cross-section, converts each velocity-pressure reading to velocity, and combines the readings according to the selected method.

Select the plane from the governing procedure

The best plane is a straight, constant-area duct segment away from elbows, transitions, dampers, branches, fan discharge, and other disturbances. The required upstream and downstream clearances depend on the applicable standard or method and the disturbance. A commonly encountered ideal criterion is 7.5 equivalent diameters downstream and 2.5 diameters upstream, but it must not be presented as the minimum for every procedure or field condition.

For the hydraulic-diameter calculation used by many TAB procedures for rectangular duct:

De = 2WH / (W + H)

For a 30-inch by 20-inch duct, De = (2 × 30 × 20) / 50 = 24 inches, or 2 feet. If the selected procedure calls for 7.5 De downstream, that distance is 15 feet. This calculation applies the stated criterion; it does not create one where the governing procedure specifies something else.

Inspect the candidate plane. Avoid a location with visible swirl, pulsation, stratification, loose liner, an obstructed probe path, or leakage at test holes. Confirm internal dimensions rather than relying only on nominal drawing size. Record the actual distances to disturbances and the reason for choosing the plane. When an ideal plane is unavailable, use the best permitted location, increase diagnostic checks as appropriate, compare with an independent method, and disclose the limitation.

Rectangular-duct layouts

Point coordinates come from the named traverse method. Log-Tchebycheff layouts place more points near the walls so the sample represents boundary-layer effects. Equal-area layouts place each reading at the representative location of an equal-area cell. Either label is incomplete without the actual coordinate table and required number of points from the governing procedure.

Do not invent coordinates in the field. Calculate each insertion depth from the measured internal dimension and the published fractional location. Mark the probe stem, keep it normal to the duct wall, and account for liner thickness and wall offset. A point accidentally placed against a wall, behind a turning vane, or outside the intended cell is not a valid sample.

Round-duct layouts

Round-duct procedures commonly use readings along two perpendicular diameters so both directions of the velocity field are represented. The number and radial positions depend on the selected equal-area or Log-T method and duct size. Follow the applicable coordinate table rather than assuming that one fixed count—such as six, eight, or ten points per diameter—always governs.

Use access holes that allow the probe to reach each required position without contacting the opposite wall. Align the impact opening into the airstream. If yaw or pitch is suspected, rotate and compare the probe cautiously under the instrument procedure; severe directional flow is also evidence that the location may be poor.

Field execution

Before collecting data, verify system mode and stability, inspect and zero the manometer, leak-check tubing, and confirm that the pressure range can resolve the expected signal. At every point:

1. place the probe at the calculated coordinate;
2. align it with the local flow as the method permits;
3. allow the reading to stabilize;
4. record the signed velocity pressure and point identity; and
5. investigate negative, erratic, or implausible values rather than changing them.

Seal test holes after the traverse. A defensible report identifies duct dimensions, net area, point method, plane location, disturbances, instrument, individual readings, air-density treatment, average velocity, and calculated airflow.

Quality checks

Convert velocity pressure to velocity before averaging because velocity varies with the square root of velocity pressure. Compare the traverse total with fan inlet or outlet readings, terminal sums, an airflow station, or another suitable method. The readings need not match exactly, but a material difference requires investigation of leakage, diversity, density, simultaneous operating conditions, instrument setup, or poor profile. Repeating a traverse in the same bad plane does not remove the underlying limitation.

Location-versus-layout distinction

A correct point grid cannot cure a severely disturbed plane, and a long straight run cannot cure wrong point coordinates. Evaluate these independently. The report should allow another technician to locate the plane, reconstruct every insertion depth, and understand any compromise.