Current research in the density area is aimed at providing traceable measurements for the internal volume of artefacts of non-standard shape.

A traditional and widely-used method for the determination of internal volume is that of pyknometry. The artefact under test is weighed both empty and then full of a liquid. Knowing the change in mass and the density of the fluid the volume can be calculated. The major uncertainty contributions arise from the weighing process, knowledge of how full of liquid the artefact is, and how accurately the density of that liquid is known.

Weighing the artefact and determining the liquid density are achievable within the required uncertainty. This leaves the filling method, particularly for non-standard artefacts, as the major problem. It is very easy to leave trapped air within the artefact, which would lead to inaccurate and non-repeatable volume determinations.

To provide confidence in the results two methods are being investigated. The first is the filling of the artefact under vacuum, using a pure hydrocarbon. The second is to fill with water by boiling it into the artefact, thereby removing all air from the water.

A major driver for this research is the NPL Boltzmann project, where the determination of the internal volume of a reference volume is required with an uncertainty of a few parts per million. However, the method once developed can be used for the volume determination of a wide range of other artefacts.

Other research currently underway at NPL includes the determination of the density of very small samples (<1 gram) to uncertainties better than 1 part per million.