Skin Friction Measurements of Systematically-Varied Roughness: Probing the Role of Roughness Amplitude and Skewness

Skin-friction, roughness functions and predictive correlations are presented for random roughness that has a Gaussian power spectral density distribution of surface elevations. The root-mean-square (rms) roughness height and the skewness of the probability density function are parametrically varied to investigate the role of these parameters in generating the friction at the wall. Results are presented for all roughness regimes, from hydraulically-smooth to fully-rough. Negative skewness (pits) had a much smaller influence on drag than positive skewness (peaks). Predictive engineering correlations for the equivalent sandgrain roughness height indicate that the rms roughess height and skewness are important scaling parameters. However, the scaling does not appear to be universal as different correlations are needed for surface roughness with positive, negative and zero skewness. Most surfaces collapse to a single roughness function in the transitionally-rough regime similar to the one developed by Nikuradase (1933) for uniform sand-grain roughness. The exceptions are the wavy surface (low effective slope) and the surface with high positive skewness.