(43e) Investigating the Effect of Pipe Diameter Contraction on the Breakage of Whey Protein Aggregates Using Cfd

Some industrial pipelines have a sharp pipe diameter reduction from a wide pipe to a narrow one, creating a characteristic flow known as converging flow. Some researchers have studied the breakage of aggregates and flocs under converging flow and concluded that particle breakage occurs just before the contraction though not after it. In the present work we studied the breakage of whey protein aggregates in a sharp pipe diameter contraction and in a smoother one; commercial CFD code was used to model the flow field in both contraction types at different bulk fluid velocities. The CFD simulation results coupled with the experimental data concluded that the pipe contraction type has a strong influence on the particle breakage and the particle size reduction occurs along the pipe and not only in the contraction region. Whey protein aggregates, formed in a standard configuration stirred tank reactor at average shear rates of 357 s^-1 and 1118 s^-1, were used as a case study in the present work. The dispersion (originally 10% solids by weight) was then diluted to less than 0.1% solids content to negate any particle-particle interaction in the breakage experiments. The diluted dispersion was propelled through a pipeline at mass flow rates between 2.4 to 7.1 gs^-1 using compressed air at different pressures. The dispersion was passed 50 times through the pipeline to study the effect of the exposure time on the particle breakage. The pipeline consisted of 55mm of wide pipe (ID 2.286mm) and 100mm of narrow pipe (ID 0.736mm), two different pipe diameter contractions were studied: the sharp contraction and the smoother one. The narrow pipe was drilled at 45 °to create a smoother transition. The pipelines including sharp and smoother transitions were modelled using a commercial CFD code; the standard κ-ε model was used during the simulation. The experimental results show that at the same mass flow rate more particle breakage occurs for the sharp transition than for the smoother one and that the particle size reduction is a function of both the mass flow rate and the exposure time. The CFD simulation results show that at the same mass flow rate, say 7.1 gs^-1, there are important variations in the flow field between sharp and smooth transition, e.g., the turbulence eddy dissipation rate (ε) for sharp transition has a maximum of 183 000 m²s^-3 and average of 21 000 m²s^-3, while for the smoother transition the maximum is 39 000 m²s^-3 and the average is 20 000 m²s^-3. It was observed that for sharp transition there is a region with very high eddy dissipation rate that starts just before the contraction and remains for 5 mm after it, while for smooth transition this region is practically avoided as one can conclude from analysing the turbulence eddy dissipation maximum and average figures presented above. This difference in ε can lead to more particle breakage and it was used to explain our experimental results. A predictive model of the form: Δd= k*Δt*(ε_i -ε_th)/(ε_th)ⁿ was used to describe the particle breakage. It relates the particle diameter reduction (Δd) to the exposure time (Δt) to a local turbulent eddy dissipation rate (ε_i) and the maximum dissipation rate (ε_th) that a particle with a given size can withstand before any breakage occurs. The adjustable parameters were fitted using the experimental data obtained from the smoother transition and then this model was validated predicting the particle size for the sharp transition flow geometry at different mass flow rates. The latter prediction proved to be in good agreement with experimental data.