Fluid property tables
Every estimate in the Fluid Dynamics tab is driven by four numbers per fluid. These are the values OpenChip ships. All are quoted at 20 °C unless stated.
The presets
Section titled “The presets”| Fluid | Density (kg/m³) | Viscosity (mPa·s) | k (W/m·K) | cₚ (J/kg·K) |
|---|---|---|---|---|
| Water | 998.2 | 1.002 | 0.598 | 4182 |
| PBS buffer (1×) | 1005 | 1.02 | 0.600 | 4100 |
| Ethanol | 789 | 1.20 | 0.171 | 2440 |
| Mineral oil (light) | 850 | 30 | 0.140 | 1900 |
| Glycerol/water 50 % | computed | computed | computed | computed |
| Blood-mimicking analog | computed | computed | computed | computed |
Viscosity is stored in Pa·s and shown here in mPa·s, which is what most tables and most bottles use. 1 mPa·s = 0.001 Pa·s.
What the caveats mean
Section titled “What the caveats mean”PBS buffer. Close enough to water for most purposes. The dissolved salts change density more than they change viscosity.
Mineral oil. Grade-dependent, and the spread is large: light oils run roughly 15–40 mPa·s and heavy ones far more. The shipped 30 mPa·s is a mid-light oil. Check your bottle — this is the one preset where using the default without looking will give you a pressure estimate that is wrong by a factor of two or more.
Glycerol/water blends. Computed rather than tabulated, from the mass fraction and temperature. The relationship between glycerol percentage and viscosity is strongly non-linear — 50 % is about six times water, 80 % is about sixty times — so the app exposes a percentage control rather than making you look up a number.
Blood-mimicking analog. A Newtonian glycerol/water blend at 40 %, matched to whole blood’s high-shear viscosity.
Whole blood is shear-thinning: its apparent viscosity falls from around 10 mPa·s at low shear to about 3.5 mPa·s above roughly 100 s⁻¹. Most microchannels at working flow rates sit in the high-shear regime, which is the one this analog matches. At very low shear rates it will under-predict the real thing. If your design has a slow reservoir or a stagnation region, the numbers there are optimistic.
This is the only preset that carries a prominent non-Newtonian note in the app, and it carries it because the whole flow model assumes a Newtonian fluid.
Adding your own
Section titled “Adding your own”Fluid Dynamics → the fluid list → New fluid. You get all four fields plus a free-text note. A custom fluid is stored in the project, not in your settings, so it travels with the file.
Seed values come from water, so every field starts plausible rather than empty.
The four numbers you need:
| Field | Where to find it |
|---|---|
| Density | Supplier data sheet, or weigh a known volume. |
| Dynamic viscosity | Data sheet, or a viscometer. This is the one that matters most for pressure. |
| Thermal conductivity | Only used for heat-exchange estimates. Water’s value is a reasonable stand-in for most aqueous solutions. |
| Specific heat capacity | Same — only used for heat exchange. |
If you are not doing heat exchange, the last two do not affect any number you will look at. Leave them at water’s values.
Where these values stop being trustworthy
Section titled “Where these values stop being trustworthy”- Temperature. Everything here is at 20 °C except the glycerol blends, which are computed at whatever temperature you set. Water’s viscosity roughly halves between 20 °C and 60 °C, so a heated run at the default value over-predicts pressure substantially.
- Non-Newtonian fluids. Polymer solutions, cell suspensions, whole blood at low shear, anything with particles. The model has one viscosity number per fluid and no shear-rate dependence at all.
- Suspensions and emulsions. An effective bulk viscosity is a rough approximation, and it stops being one at all once particle size approaches channel width.
- Surfactants. Not modelled. They change interfacial behaviour, which is most of what a droplet generator depends on.
See The limits of analytical estimates for the equivalent statement about the flow model itself.