Comments on ” Microscale flow visualization ”

We make comments on the presentation of Sinton’s paper (Microfluidics and Nanofluidics 1: 2, 2004) about the microscale flow visualization since the effects of the roughness along the microfabricated wall upon the current macroflow visualization methods could be significant and cannot be neglected in microdomain and even nanodomain. Sinton just presented a rather interesting and comprehensive review about the microscale flow visualization (Sinton, 2004) due to advances in microfluidic and nanofluidic technologies (Li, 2004) being paralleled by advances in methods for direct optical measurement of transport phenomena on these scales. As Sinton noticed, a variety of methods for microscale flow visualization have appeared and evolved since the late 1990s. These methods and their applications to date are reviewed therein (Sinton, 2004) in detail, and in context of the both the fundamental phenomena they exploit and the fundamental phenomena they are applied to measure. Where possible, links to macroflow visualization methods are established, and the physical mechanisms underlying these methods are explained. We all know that direct flow visualization is of key importance for the fundamental understanding of microflows, analyzing, developing and evaluating novel microfluidic processes, investigating non-ideal behavior such as spatial and temporal gradients in surface and fluid properties, and providing benchmark data for computational investigations (Sinton and Li, 2004). There are, however, some differences between macroand microdomain which will influence the results based on particle-based, scalar-based and point-detection scanning microfluidic flow visualization methods. One specific example is the roughness of the bounded wall, which is quite significant in common microchannels but could be ad hoc neglected in macrochannels (Chu, 2000/2002). For microfabricated Si-based walls, the roughness cannot be eliminated completely and thus induce quite random scattering effects (e.g., for the particle image velocimetry (PIV), the motion of the bulk fluid is inferred from the observed velocity of marker particles, Adrian 1991) especially for the particle-based flow visualization methods. As claimed before (Santiago et al., 1998), the stochastic influence of the Brownian motion of the small particles was significant, however, ensemble averaging over several images was shown to greatly improve the obtained velocity field. It is clear that the near-wall flow field would have poor resolution due to the irregular scattering coming from the random roughness along the wall or the confined boundary (Chu, 2000/2002). How about the scalar-based flow velocimetry, where the motion of the bulk fluid is inferred from ∗Present Address : P.O. Box 39, Tou-Di-Ban, Xihong Road, Urumqi 830000, PR China.