A Source of Systematic Error in Transient‐Flow Air Permeameter Measurements

Abstract The data from transient‐flow air permeameters often exhibit curvature in the theoretically linear plots of the natual log of pressure vs. time. This was originally attributed to errors in water manometer data caused by the inertia of the water in the manometer. The resulting recommendation was to ignore the early time data when calculating the air permeability of a soil by these methods. We have demonstrated and quantified that the exhibited curvature is a direct consequence of small changes in temperature in the source air tank as a result of the cooling of the expanding air during the permeability determination. This was accomplished by interfacing the air permeameter with a computer to collect pressure and temperature data at a relatively high frequency. The data acquisition procedure facilitated the calculation of the change in pressure and temperature with respect to time, which allowed the direct solution of the differential form of the equations describing the mass flux of air from the tank through the core. The data is compared with the theoretical relationship expected for the core based on air permeability determined using a steady‐state method.