Viscometer and rheometer calibration: temperature, shear rate and standards

Two errors dominate viscosity work and neither is on the certificate. The first is temperature: for many liquids viscosity changes by around 2 % per degree near room temperature, and for some oils considerably more, so a bath running half a degree high can shift the answer by more than the instrument's entire measurement uncertainty. The second is reporting one number for a fluid whose viscosity depends on how hard you shear it, which covers most suspensions, gels, polymer solutions and semi-solid formulations.

Both are method problems rather than instrument problems, which is why a laboratory with an immaculately calibrated rheometer can still produce viscosity data nobody can use.

Temperature is the dominant error

The standards treat temperature as a primary control, not an environmental note. A capillary viscosity determination is quoted at a stated temperature because the number has no meaning without it, and the bath has to be both stable and uniform — stable so the measurement does not drift during the flow time, uniform so the whole capillary is at the stated temperature rather than the average of a gradient.

• Calibrate the bath's own thermometer, and calibrate it at the temperature you work at rather than at one convenient point.
• Measure the gradient across the working volume, not just the control point. A bath whose controller reads 40.00 °C can be 39.8 °C at the far end.
• Allow genuine equilibration. A viscometer dropped into a bath is at bath temperature on the outside long before the sample inside it is.
• For a rotational instrument, remember that shearing a viscous sample heats it. At high shear the sample can be meaningfully hotter than the plate.
• Record the temperature with every result, to the resolution that matters for the fluid.

Capillary viscometers: the constant is the calibration

A glass capillary viscometer measures the time for a fixed volume of liquid to flow under gravity through a capillary. Kinematic viscosity is that time multiplied by the viscometer's constant, so the constant carries the entire traceability of the measurement. ISO 3105 specifies the viscometers used for this and describes their calibration; the types it covers are modified Ostwald, suspended-level and reverse-flow designs. ISO 3104:2023 is the determination itself, with a Procedure A using manual glass viscometers and a Procedure B using them in an automated assembly, over a range from 0.2 mm²/s to 300 000 mm²/s and temperatures from −20 °C to +150 °C. ASTM D445 is the corresponding ASTM method.

The failure mode practitioners actually hit here is a slow upward creep in results over months, across all fluids, on one viscometer. It is almost never the fluid and almost never the bath; it is deposit inside the capillary. Because the change is gradual and affects everything equally, it survives a comparison between samples and only becomes visible against a certified reference fluid — which is the argument for running one periodically rather than only at calibration.

Rotational instruments: torque, speed and geometry

A rotational viscometer or rheometer measures the torque needed to turn a defined geometry at a defined speed in the sample. Three things therefore have to be right: the torque transducer, the rotational speed, and the dimensions of the measuring system. ISO 3219-1:2021 sets the vocabulary and symbols for rotational and oscillatory rheometry, replacing the 1993 edition of ISO 3219; DIN 53019 covers measurement with rotational viscometers, with Part 1 on principles and geometry and Part 2 on calibration and the determination of measurement uncertainty for coaxial cylinder and cone-and-plate systems.

Gap setting is the one that bites. On a cone-and-plate or parallel-plate instrument the shear rate is calculated from the gap, so a zero-gap determination that is 20 µm out puts a proportional error into every shear rate and therefore every viscosity, with no indication on screen. Thermal expansion of the shaft and plate during warm-up is enough to do it, which is why the gap is zeroed at the measurement temperature after the instrument has equilibrated.

Newtonian and non-Newtonian: why a bare number fails

For a Newtonian fluid, viscosity is independent of shear rate and a single value genuinely describes it. For a shear-thinning fluid — most suspensions, emulsions, polymer solutions, gels and creams — the measured viscosity falls as shear rate rises, and a result reported without its shear rate cannot be reproduced even by the same laboratory on the same instrument.

ISO 3104 makes the related point about capillary work: if viscosity varies significantly with shear rate, viscometers with different capillary diameters can give different results, because they shear the sample differently. The fluid is not misbehaving; the measurement is being asked a question that has more than one answer.

For any non-Newtonian material the reportable result is a viscosity at a stated shear rate and temperature, or a flow curve. If a specification states a single viscosity figure without a shear rate, the specification is incomplete and the measurement will be argued about later.

Time dependence adds a second dimension. Thixotropic materials recover structure at rest and lose it under shear, so the result depends on the shear history: how long the sample sat after loading, whether it was pre-sheared, how fast the ramp went. These are method parameters that have to be fixed and written down, and two laboratories following the same numerical specification with different rest times will not agree.

Certified reference fluids

Traceability in practice comes from certified viscosity reference fluids: oils with a certified viscosity at stated temperatures and an uncertainty on the certificate. Choosing one is a matter of bracketing the work — a fluid near the viscosity you measure, at the temperature you measure at — and of respecting the certificate's conditions, since the certified value at 40 °C says nothing directly about behaviour at 25 °C.

• Use a fluid whose certified viscosity is near your working range; a 5 mm²/s standard tells you little about work at 5000 mm²/s.
• Honour the certified temperature, and interpolate only if the certificate supports it.
• Treat the fluid as consumable and record its lot and expiry — these oils change with exposure and contamination.
• Do not return used fluid to the bottle.
• For a rotational instrument, remember that a Newtonian reference fluid checks the torque and speed but exercises none of the instrument's behaviour on structured samples.

Before blaming the instrument

In pharmaceutical work USP <911> covers capillary viscometer methods and USP <912> rotational rheometer methods, and either will be the method of record where a monograph calls for viscosity. If results disagree with expectation, the order of investigation that resolves most cases is: temperature at the sample, not at the controller; shear rate and whether it is the same one the specification meant; sample preparation and rest time; geometry, gap and fill; and only then the calibration of the torque and speed. Dynamic viscosity is kinematic viscosity multiplied by density, so where the two are converted, the density figure is part of the measurement and a density meter result carries its own uncertainty into the answer.

ValiTrac AI is not an accredited calibration laboratory and confers no compliance. Viscosity specifications and the acceptance criteria for any check described here belong to your own method, monograph or quality system.