Objective:

To present a methodology that allows experimentally determining shear velocity, considering the log-law as a model of velocity distribution in the outer region of turbulent flow.

Mathodology:

The experimental study was carried out in a rectangular-shaped, variably-sloped channel with a 0.245-m-wide base and 5 m long. Flow velocity was measured with an Acoustic Doppler Velocimeter (ADV), and the measurement area was 12 mm. Shear velocity was determined by the instantaneous velocity equation ($u_{i,j}$).

Results:

The log-law model had a good statistical fit with the shear velocity estimated from the experimental data.

Study limitations:

The experimental tests were conducted only in subcritical regime with low aspect ratios. In addition, in all tests, the measurement of instantaneous velocities was carried out only in a 12-mm profile, as close as possible to the wall.

Originality:

The model to calculate the shear velocity is presented explicitly, and the statistical

approach employed supports the use of the median as an estimator of the shear velocity.

Conclusions:

The presented methodology shows low uncertainty in the estimation of shear velocity. The Anderson-Darling test showed that the results do not follow a normal distribution, so the median is the statistical parameter to define the shear velocity value.