We are often asked how fast fluids move through microfluidic channels. The answer is a function of several factors, including the liquid viscosity, the pressure applied on the fluid, and the microfluidic channel geometry. Using one of our fused silica glass microfluidic chips, we collected some experimental data, and we compared it to the prediction of the Hagen–Poiseuille theoretical model.

Volumetric flow rates as a function of pressure and viscosity: Experimental data

As mentioned above the flow rate is a function of the pressure applied onto the fluid, or more exactly, a function of the pressure differential between the input and the output. In all of our measurements, the output was free flowing (no back pressure present beyond atmospheric pressure). We ran our tests using pressures ranging from 3 psi to 15 psi (~20 to 100 kPa). Our microfluidic set-up was powered using a pressurized air reservoir. 

The flow rate is also a function of the viscosity of the fluid. In order to get a wide range of viscosities, we used three fluids: Deionized Water, which has a rather low viscosity. Hexadecane, which is about thirty-times more viscous than water, and canola oil which is even more viscous (one order of magnitude more viscous than hexadecane).

For the initial test, we used a very simple straight-channel flow cell (three-layer microfluidic chip) with the following dimensions:

As evident from the accompanying experimental data graph, the flow rates measured are proportional to the applied pressure and inversely proportional to the fluid viscosity.

Can this be modelled? Yes, rather well in fact. Learn more in this follow-up article.