Hide/Show Apps

Simulation of laminar microchannel flows with realistic 3D surface roughness

Akbaş, Batuhan
Effects of flow development and surface roughness on the pressure drop characteristics of laminar liquid flowsinside microchannels are investigated numerically using OpenFOAM. Channels with square cross section of 500 μm×500 μmand length of 80 mm are studied. Top surface of the channels are artificially roughened using thespatial frequency methodto create 8 different roughness profiles. Scaling the relative roughness ({u1D700}) values of each profile to three different values (1.0, 2.5and5.0 %),a total of 24 cases are simulated in the Reynolds number range of 100-1500. Results are presented mainly as the variation of the apparent Poiseuille number (Poapp) with Reynolds number (Re). Flow fields are examined by considering velocity magnitude contours at different cross sections to understand the flow development effects, as well as streamlines. For 1% relative roughness, it is seen that {u1D700}is a representative parameter to understand the pressure drop behavior and other roughness characteristics do not affect the flow field considerably. All {u1D700}=1% cases show Poapp–Re variations that have the same behavior as that of the smooth channel theory, with maximum deviations being 3% at Re = 100 and 7% at Re = 1500. However, for {u1D700}=2.5% and 5% cases, {u1D700}is no longer enough to explain the effect of surface roughness on pressure drop. Although certain roughness profiles provide Poapp–Re variations that are similar to that of the smooth channel theory, other profiles resulted in completely different behaviors with deviations increasing with Re. It is seen that for {u1D700}=2.5% Re = 1500, depending on the surface profile, Poappcan deviate from the smooth theory as low as 15% and as high as 55%. These numbers increased to 37% and 151% for {u1D700}=5% cases.It is concluded that the misunderstandings and the conflicting results seen in the literature about pressure drop in microchannels can be related to the use of relative surface roughness as the only defining parameter of roughness or even completely omitting roughness effects, which is clearly not the case for relative surface roughness values higher than 1%. Other parameters such as number of peaks and waviness also need to be considered and reported.