Flow characteristics of Newtonian and non-Newtonian fluids in a channel with obstruction at the entry

This study investigates the flow phenomena in a channel with an obstruction at the entry which is placed in another wider parallel walled channel. When obstructed, the flow phenomena inside the channel were observed to be reverse, forward or stagnant depending on the position of the obstruction. The parameters that influence the flow inside and around the test channel are: - the size and shape of the obstruction geometries, the gap between the test channel and the obstruction geometry, the Reynolds number and the length of the test channel. Knowledge of these flow phenomena has the potential benefit in the control of various flows in process engineering applications.

Experimental investigations of these flow parameters were carried out in an open channel rig. Fluids used in the investigations were a Newtonian fluid (water) and two non-Newtonian fluids, namely polyacrylamide solution (0.03% by weight) and mixed solution (xanthan gum, magna floc 139 and magna floc 1011). The polyacrylamide solution and mixed solution had similar viscosity and both show a power-law behavior, however their elastic behavior was different.

Experimental studies of these flows include the velocity measurement and the flow visualization analysis. The velocity measurement provides the quantitative information whereas flow visualization provides the qualitative information of the flow. Numerical simulations of these flow phenomena were also carried out using a CFD software and comparisons are made with the experimental results.

The influence of the size and shapes of the obstruction geometries; and the gap to width (g/w) ratio on the magnitude of the velocity ratio (V_iN_o: inside/outside velocity of the test channel) was studied. Obstruction geometries used were semicircle, triangle, circle and various shapes of rectangles. The g/w ratios ranging from 0.5 to 8 were selected as a set of distances from the test channel. The influence of the Reynolds numbers on the value of the velocity ratio was investigated. The effect of the test channel length on the velocity ratio was also investigated at the Reynolds number of 2000 for the above specified g/w ratios.

The flow inside the test channel was observed to be forward, reverse or stagnant for both Newtonian fluid (water) and Non-Newtonian fluids. The 'flat plate' obstruction geometry produced the maximum reverse flow inside the test channel compared with other obstruction geometries for both Newtonian and non-Newtonian fluids. The magnitude of the reverse flow for both non-Newtonian fluids used in this study is observed to be half of the magnitude of the reverse flow for water. The maximum reverse flow for non-Newtonian fluids occurs at g/w ratio of 1.0 whereas for Newtonian fluid (water) it occurs at g/w ratio of 1.5.)

The two flow parameters namely, the size and shapes of the obstruction geometries and the gap between the test channel and the obstruction geometries have the strongest influence on the flow phenomena. The Reynolds number has also a strong influence whereas the test channel length has a negligible influence on the flow phenomena.

The numerical simulations using CFD-ACE+ found that the numerically predicted streamlines and velocity vectors of the flow phenomena are in good agreement with the streak lines of the flow visualization images. It was also found that the numerical model used for this study can be generally applied for the prediction of the flow behaviour in the channel with obstruction at the entry.