(522d) Approach towards a Quantitative Description of Microstructured Cyclone Type Mixers

Approach Towards a Quantitative Description of Microstructured Cyclone Type Mixers

Authors

A. Kölbl*, M. Kraut, A. Wenka, and R. Dittmeyer:

Institute for Micro Process Engineering, Karlsruhe Institute of Technology, 76344 Eggenstein-Leopoldshafen, Germany.

*Telephone: 07247/82-6653, Fax 07247/82-3186, Email: Andreas.Koelbl@kit.edu

Mixing of two liquids in miniaturized Cyclone Type Mixers (figure 1) is achieved by feeding them tangentially into a mixing chamber.

Figure 1: Principal sketch of Cyclone Type Mixers.

Thus a swirl motion is created, which enhances mixing. We recently published a parameter study on the qualitative influence of various structural parameters on mixing in such devices [Kölbl2010] based on computational fluid dynamics (CFD) and experiments using the Iodide Iodate Reaction Method. It was found that the diameter of the cyclone mixing chamber is more important for mixing than the mixing chamber length. Furthermore, the results from numerical simulations indicate a swirl motion in the mixing chamber as well as the occurrence of cavitation and zones of complete recirculation (dead zones).

In this contribution we develop a quantitative description of mixing in Cyclone Type Mixers. The influence of the mass flow throughput on the 'mixing quality' is discussed as well as the influence of the viscosity on hydrodynamics, pressure drop, and mixing.

Influence of the mass flow rate on mixing

The results from CFD calculations (Fluent 12.1) indicate that the distribution of a passive inert tracer is almost independent on the mass flow rate in the mixing device. Figure 2 shows the normalized distribution of a passive inert tracer at two different mass flow rates: (a) 2x1.5 kg/h and (b) 2x3 kg/h.

Figure 2: Distribution of a passive inert tracer at two different mass flow rates: (a) 2x1.5 kg/h and (b) 2x3 kg/h.

The findings suggest that the mean mixing time, defined here after Lindenberg et al. [Lindenberg2008] as the residence time along the stream lines until complete mixing is reached can be directly adjusted by the mass flow rate. An experimental verification of this scaling assumption is provided by applying the Iodide Iodate Reaction Method. Assuming that the triiodide concentrations in the resulting mixtures are directly proportional to the mixing times [Falk2010], the triiodide concentrations should be proportional to (dm/dt)^-1. This assumption is verified through the observation that the triiodide concentrations in the resulting mixtures can be described by a simple empirical correlation (equation 1) for various devices and concentration sets.

(1) c(triiodide) = A/(dm/dt)

Thus the triodide concentrations only depend on a single parameter A, which is a function of the reactant concentrations, the viscosity, and geometrical details of the examined mixing devices.

Influence of the viscosity on hydrodynamics, pressure drop, and mixing

Results from numerical simulations further indicate that the pressure drop in the mixing chamber, which is decisive for the mixing results is independent of the viscosity. It should be kept in mind that the pressure drop in the mixing chamber which includes feed and exit channels and fluid bends accounts only for about 50 % of the whole device [Kölbl2010]. However, the number of times the fluids rotates in the mixing chamber is reduced at higher viscosities. Thus the contact area between the fluids is diminished considerably. Figure 3 shows path lines for a fluid with a viscosity of (a) 0.5 mPas and (b) 10 mPas in the same device. The entry velocity is 5 m/s in both cases.

Figure 3: Pathlines for fluids of different viscosities: (a) 0.5 mPas and (b) 10 mPas.

Due to the reduction of fluid turns in the mixing chamber, reduced mixing qualities are expected for fluids at elevated viscosities. This expectation is verified by applying the Iodide Iodate Reaction Method for viscosities of 1 mPas (aqueous solution) and 2 mPas (aqueous solution + glycerol). For Cyclone Type Mixers the ratio of the optical densities of the resulting solutions obtained at 2 mPas and 1 mPas is independent of the flow rates applied. This is another experimental finding that supports the sacling assumption from CFD calculations. However the influence of the elevated viscosity is rather small compared to microstructured V-type mixers [Kölbl2009].

[Kölbl2010]

A. Kölbl, M. Kraut, and A. Wenka:

?Experimental Parameter Study on Cyclone Type Mixers?

11^th Conference on Micro Reaction Technology (IMRET 11), Kyoto, Japan, March 8.-10. 2010.

[Lindenberg2008]

C. Lindenberg, J. Schöll, L. Vicum, M. Mazotti, and J. Brozio:

?Experimental characterization and multi-scale modeling of mixing in static mixers?

Chemical Engineering Science 63(2008)4135-4149.

[Falk2010]

L. Falk and J.-M. Commenge:

?Performance comparison of micromixers?

Chemical Engineering Science 65 (2010) 405-411.

[Kölbl2009]

A. Kölbl, M. Kraut und K. Schubert:

?Mischexperimente mit einem Mikromischer bei verschiedenen Viskositäten?

Chemie Ingenieur Technik 81(2009)1058-1059.