Pressure hole errors in the annular flow measurement of the second normal stress difference

Abstract

There is a critical need to accurately determine viscoelastic fluid material functions, one of which is the second normal stress difference. While it is now generally accepted that this function is non-zero, in contradiction to the Weissenberg hypothesis, numerous investigators have reported values of opposite sign as well as significant unexplained differences in magnitude. Ginn and Metzner (3) have critically examined much of the data and many techniques. They show that the sign of all internally consistent second normal stress difference data is obtained as negative by all techniques except annular flow devices, which uniformly yield positive values. They also note that, even with considerable care, accurate second normal stress difference data are not likely to be obtained using either the parallel plate or cone and plate systems, especially at higher shear rates. The more recent work of Miller and Christiansen (10) seems to improve the accuracy of measurement of the second normal stress difference using both the rim-pressure and integral-force methods of obtaining and analysing cone and plate data. Their results are still somewhat erratic, however, and differ by up to a factor of two. In any event, inertial effects preclude the use of these techniques much above shear rates of 500 sec⁻¹.