The publication of a theoretical analysis of the response of a simple strai
ght-tube Coriolis meter to flow pulsations raised the question of the exten
t to which the results of that analysis are generic over the wide range of
geometric configurations used in commercially available meters. A procedure
for using a general purpose finite element (FE) code to investigate this q
uestion is presented. The dual time scales, which are an essential feature
of pulsating how through a Coriolis meter, are used to minimize the amount
of computation required to simulate the meter response. The FE model is dev
eloped in a full 3-D form with shear deflection and axial forces, and the c
omputation of the simulated response for the geometrically most complex met
er currently available shows that this level of representation is necessary
to reveal the full details of the response. The response derived from the
FE simulation for straight-tube meters, is compared with the published theo
retical response and to experimental data. Over a range of different meters
, the characteristics of the sensor signals in the presence of flow pulsati
ons are shown to be generally similar. In all cases, the simulated sensor s
ignals contain components corresponding to beating between the pulsation fr
equency and the meter drive frequency, in addition to the main component at
the drive frequency. Spectra are computed from the simulated meter respons
es and these are used to show that the relationship between the mass flow r
ate and the phase difference between the component of the sensor signals at
the drive frequency, is not significantly affected by the pulsations. Thus
, the work suggests that the reports of changes in meter calibration due to
certain frequencies of flow pulsation represent errors in signal processin
g rather than fundamental changes in the meter characteristics. (C) 2000 Ac
ademic Press.