Fluid properties for pipe flow calculations

Kinematic viscosity is given in mm^2/s, which is the same unit as the centistoke, and dynamic viscosity in mPa.s, which is the same as the centipoise. The two are related by mu = nu * rho, and the dynamic column is computed from the other two rather than tabulated separately, so it cannot disagree with them.

Water is a special case, in a good way. The water rows are not frozen numbers. Density and viscosity are evaluated from the solver's own temperature correlations at the stated temperature when this page is built, and the specific heat and thermal conductivity are interpolated from tabulated water data at ten-degree intervals between 0 and 100 degrees C, taken from Incropera and DeWitt, Fundamentals of Heat and Mass Transfer, Table A.6. In the Studio, a water fluid keeps following those correlations as the temperature changes through a thermal solve, rather than sitting at whatever the panel was set to.

Vapour pressure. The water rows show a dash in that column on purpose. Rather than freeze one number, the solver evaluates the Antoine correlation for water at the local temperature wherever it needs a vapour pressure, which is the right behaviour because a fixed value would be wrong everywhere except at the reference temperature. Every non-water preset carries an explicit vapour pressure, which is what switches the NPSH available and flashing diagnostics on for it. The oils carry zero, which is not laziness: they are effectively non-volatile at these temperatures and zero is the honest entry.

Where the numbers come from. Water follows the correlations and tabulations named above. The remaining liquids are representative published values for the named fluid, grade or concentration at the stated temperature. Petrol, crude oil, sodium hydroxide solution and the propylene glycol mixtures are the least certain rows: petrol vapour pressure is a seasonal specification that ranges over roughly a factor of two, crude oil varies enormously by field, and the caustic properties move sharply with both temperature and concentration. Where your answer turns on one of these, get the datasheet.

Reading the liquid table

Which viscosity does the solver want? Kinematic. The incompressible kernel works in kinematic viscosity because that is what the Reynolds number needs. If your datasheet gives dynamic viscosity, divide by the density at the same temperature.

Viscosity is the temperature-sensitive one. Density moves by a few per cent over a wide temperature range for most liquids. Viscosity can move by an order of magnitude. The hydraulic oil and gear oil rows are quoted at 40 degrees C because that is the temperature at which the ISO VG grade is defined, and the same oil at start-up on a cold morning is a different fluid hydraulically. If your system runs at a temperature far from the row you are using, correct the viscosity before you do anything else.

Vapour pressure decides pump suction, not pipe friction. It does not appear in a head-loss calculation at all. It appears in the NPSH available, where it is subtracted from the absolute pressure at the pump suction. This is why the petrol row is worth looking at even if you never pump petrol: at 60 kPa its vapour pressure is more than twenty times water's at the same temperature, and a suction arrangement that is comfortable for water is marginal for petrol.

Gases

Gas properties are given near 15 to 25 degrees C. The compressible solver takes density from the gas law rather than from a table, so what it needs is the specific gas constant, the viscosity and, for anything thermal, the ratio of specific heats and the specific heat.

The molar mass column is computed back from the tabulated specific gas constant as M = R_u / R, with R_u = 8314.462618 J/(kmol.K). It is shown because it is the number most people recognise and check against, but note what that derivation means: the last digit reflects the rounding carried in R, not a different molar mass. An automated test cross-checks each pure gas's R against the universal constant divided by its literature molar mass, so the two agree to within a tenth of a per cent.

The two mixtures are marked. Natural gas and biogas do not have a molar mass in the same sense as a pure substance, and their tabulated values are composition-weighted pseudo-properties for a typical mix. Pipeline gas composition varies by field, by season and by supplier, so for a real natural gas job enter your own gas analysis rather than using the preset. It is there to get you started, not to represent your gas.

Real-gas and viscosity-model constants

These are used only when the corresponding option is switched on. The Peng-Robinson critical constants feed the real-gas density option, and the Sutherland constant feeds temperature-dependent viscosity in a thermal solve. With both switched off, the solver uses a constant compressibility factor and a constant viscosity, which is the right choice for most low-pressure work.

Two honest cautions on the real-gas column. Peng-Robinson is a poor equation of state for hydrogen and helium, so the critical constants for those two are there for completeness and the constant-compressibility treatment is the better choice at low pressure. And for the two mixtures the critical values are pseudo-criticals for the assumed composition, which inherits the same caveat as their gas constant.

Sutherland constants are blank where the source does not give one. That is a deliberate gap rather than an oversight, and a gas without one simply keeps a constant viscosity through a thermal solve.

Non-Newtonian starting points

Fluid Network Studio solves homogeneous power-law and Bingham-plastic fluids, and it ships four example rheologies. They are not a library of fluid data and should not be read as one.

Every one of these is an order of magnitude rather than a value. Real slurries, muds and sludges vary by an order of magnitude between two samples of nominally the same material, and rheology is measured rather than looked up. The caption the app shows wherever these are offered says the same thing: measure your fluid, and note that only homogeneous, non-settling fluids are in scope. Settling slurries, deposition velocities and two-phase transport are outside what this solver models, and no rheology parameter will change that.

What these values are not

• They are not at your temperature. Each row is a single reference temperature, printed in the row. Interpolating between two of these rows is not a temperature correlation.
• They are not at your pressure. Liquid density and viscosity are given at atmospheric conditions. That is fine for the great majority of pipe work and it is not fine for high-pressure hydraulic systems.
• They are not grade-specific. "Diesel", "light oil" and "crude oil, light" are categories, not products. Two diesels from two refineries differ.
• They are editable, and should be edited. In the Studio every property a preset fills stays editable. The presets exist so a model runs before the datasheet arrives, not so the datasheet can be skipped.

Frequently asked questions

What is the density and viscosity of water at 20 degrees C?

About 998.2 kg/m^3 and a kinematic viscosity of about 1.004 mm^2/s, which is a dynamic viscosity of about 1.002 mPa.s. Those figures are in the table above, computed from the solver's own correlations rather than transcribed, so they are the exact values a solve at 20 degrees C uses.

What is the difference between kinematic and dynamic viscosity?

Dynamic viscosity is the fluid's resistance to shear, in Pa.s. Kinematic viscosity is that divided by density, in m^2/s, and it is what the Reynolds number needs. In practical units, mPa.s is the centipoise and mm^2/s is the centistoke, so a datasheet in cP or cSt is already in the units of the two viscosity columns above.

Which properties do I actually need for a pressure drop calculation?

Density and viscosity, and nothing else. Specific heat and thermal conductivity are only needed once heat transfer is switched on, and vapour pressure only for pump suction and cavitation checks. If you are sizing a line, two numbers will do.

Why do the water rows have no vapour pressure?

Because the solver computes it from temperature instead. Water is the one fluid where a proper correlation is worth having built in, so the Antoine equation is evaluated at the local temperature wherever a vapour pressure is needed. Every other preset carries a fixed value at its own reference temperature.

Can I use these gas properties for a compressed air calculation?

Yes. The air preset is the one the compressible solver uses by default, and compressed air work at ordinary pressures is well served by a constant compressibility factor and a constant viscosity. Turn on the real-gas and Sutherland options when pressures approach the critical region or when a thermal solve spans a wide temperature range.

What is the specific gas constant of air?

287.05 J/(kg.K), which corresponds to a molar mass of about 28.97 kg/kmol. The ratio of specific heats is 1.4 and the specific heat at constant pressure is 1005 J/(kg.K), both at ordinary ambient temperature. All three are in the gas table above.

Do you have properties for refrigerants and cryogens?

Only ammonia as a saturated liquid at 20 degrees C, for refrigeration line work, and the gas presets include hydrogen and helium. Full refrigerant property libraries with two-phase behaviour are outside what this solver does, since it models single-phase flow and single-component gases.

References

• Incropera, F. P. and DeWitt, D. P., Fundamentals of Heat and Mass Transfer, Table A.6, for the water specific heat and thermal conductivity knots, and for the gas thermal conductivities near 300 K.
• White, F. M., Viscous Fluid Flow, for the Sutherland constant of air.
• Kaye and Laby, Tables of Physical and Chemical Constants, for the Sutherland constant of methane.
• Crane Co., Technical Paper No. 410: Flow of Fluids Through Valves, Fittings and Pipe, for the Sutherland constant of carbon dioxide. tp410.com
• The Antoine constants for water used in the vapour pressure correlation are as tabulated in Lange's Handbook and the Dortmund collection.

Related

• Pipe fitting K factors and absolute pipe roughness values, the other two inputs a friction calculation needs.
• Reynolds number calculator, which is where the viscosity column ends up.
• Pump power calculator and the pump system design guide, where vapour pressure decides the suction.
• Compressed air system design guide, for what the gas constants are used for.
• Open the Studio to pick a fluid and solve a network with it.