| To further understand fluid dynamic characteristics in microchannels and to test the validity of macroscale theories that are commonly utilized at the microscale, an experimental investigation of water flow through a high aspect ratio rectangular microchannel is conducted. A rectangular microchannel with nominal dimensions of 500 mum in height, 6 mm in width, and 32.8 cm in length is CNC machined into an aluminum blank. The test-section is completed by attaching a cap blank to the microchannel blank. Both integral and velocimetry data are obtained over a Reynolds number range from 173 to 4830, where the Reynolds number is based upon hydraulic diameter and average velocity. Velocimetry data are obtained using molecular tagging velocimetry (MTV). Problems associated with the application of MTV for determination of mean velocity profiles are discussed.;Integral friction factor data are in agreement with macroscale theory and correlations for laminar and turbulent flow. The Poiseuille number for the experimental data is found to be 87.9 in comparison to the expected macroscale value of 86.22. The onset of transition occurs at a Reynolds number of 2750.;Laminar nondimensional velocimetry and coefficient of friction data are in agreement with macroscale theory. Transition from laminar flow, based upon a change in nondimensional velocity profile shape, occurs at a Reynolds number of 2800. This transitional Reynolds number is in excellent agreement with integral results and macroscale experimental results. Fully developed turbulent flow is found to exist at a Reynolds number of 2700, where the Reynolds number is based upon channel height.;Inner normalized mean velocity profiles scale for 20 viscous units, whereas the Reynolds stress and production of mean kinetic energy profiles do not scale on inner variables. The inner normalized mean velocity profiles become increasingly log-like through the transitional regime. The experimental trends for the inner normalized mean velocity. Reynolds stress, and production are in agreement with macroscale experimental and direct numerical simulation data.;The main conclusion obtained from this work is that at this scale, no microscale effect exists; therefore, designers can use macroscale theory and correlations to design microfluidic devices. Further, it has been demonstrated that MTV can be used to obtain an accurate representation of mean flow physics at the microscale. |
