The flows generated in vessels stirred by two Rushton impellers were investigated in two vessels of diameter (T)

No abstract

The velocity characteristics of the flows in a fully baffled vessel of diameter T = 290 mm stirred by a Rushton impeller of' diameter D = TI3 were investigated by means of laser-Doppler anemometry measurements. The effects of clearance and rotational speed on the flow patterns in the vessel were studied. It was found that at impeller clearances from the bottom of the vessel (C) around 0.2 Tthe characteristic double-loop flow pattern undergoes a transition to a single-loop one with the impeller stream direction becoming partly axial and being inclined at around 25 to 30" to the horizontal. The impeller stream inclination varied with radial distance from the impeller, as well as with angular position between blades (blade angle). Impeller speed was found to have no effect on the flow pattern or the mean velocities and turbulence levels normalized by V,iR for C/T> 0.20 or CIT 20.15. The flow structure measured with C = 0.1ST is described in detail and the implications of the data for fluid mixing in stirred vessels are discussed. luid motion is of paramount importance for the mixing F processes in stirred tanks but, notwithstanding their frequent use in process plants, the design of stirred reactor units is most often based on empirical knowledge and "rules of thumb" and may be subject to substantial approximations and uncertainties which cause financial losses estimated in the region of a billion dollars per annum in the USA alone (Tatterson, 1994). This is particularly true when mixing of multiphase systems is involved; for instance, in solid-liquid processes, suspension of the solids from the bottom of the vessel is normally required and impeller selection must reflect this. Axial flow impellers are normally used to lift solids off the vessel bottom, but the effects of impeller clearance and rotational speed on the mean flow pattern must be known a priori if efficient suspension is to be achieved. This is complicated, however, by quantitative and qualitative differences in the mean flow and turbulence fields which are sometimes observed with changes in the impeller speed andlor clearance, even with commonly-employed impellers such as the Rushton and the pitched blade turbine.Impeller Reynolds number or speed may affect the circulation pattern in stirred vessels and such flow changes have been previously reported for pitched blade turbines (PBTs) by Nouri and Whitelaw (1990) and Hockey (1990), who observed that the impeller stream with 45" and 60" PBTs changed direction with increasing Re from primarily radial to primarily axial flow at Reynolds numbers of 490 and 650, *Author to whom correspondence may be addressed. E-mail address: observed that the downward flowing jet from a PBT depends strongly on clearance and that impeller location is one of the critical factors for agitator design.

SUMMARYThis paper presents a combined experimental and computational study of the steady flow through an internal combustion engine inlet port. The port was of generic design with a straight centreline. The three-dimensional velocity and turbulence fields in the port and cylinder were simulated using a computational fluid dynamics programme. Laser sheet flow visualization and laser Doppler anemometry were also employed to investigate the flows and assess the predictions. The results show that a large-scale flow structure is created in the cylinder by the inlet jet and its interaction with the valve and cylinder walls. Both predictions and measurements show that the flow is strongly dependent on the valve lift but is not affected by the flow rate. Comparisons of the numerical predictions with the experimental data indicated that the mean flow features are accurately predicted in many parts of the flow field; some discrepancies are evident and stem primarily from the failure of the simulation to predict a small recirculation region in the port which affects the trajectory of the annular jet entering the cylinder. Calculations were also made without modelling the port shape by using simplified inlet conditions upstream of the valve seat. It was found that this approximation can provide a reasonable, albeit less accurate, description of the flow, but modelling of the port shape is necessary for accurate flow predictions.

No abstract