The flow-pattern in SSHEs is typically the result of axial flow and rotation of the scrapers. Some of the crucial parameters that define this flow are axial driving force, viscosity, length of the equipment, radial temperature profiles, location of the rotating shaft relative to the outer tube, number and positioning of the blades, and the type of the fluid (Newtonian, non-Newtonian) (Abichandani et al., 1987; Harrod, 1986). Flow in SSHEs can be characterized by the rotational Taylor number (Tar) and axial Reynolds number (Rea). Tar is a function of rotational Reynolds number, Rer, shaft diameter, and the annular gap for product flow.
In the absence of scraper blades and for low values of Tar, the flow is simple shear flow—often called Couette flow. In Couette flow, the axial and radial velocity components are zero.
When the rotational component of flow is gradually increased, a critical value of (Tar,Cr) is reached at which Taylor vortices start forming (Trommelen and Beek, 1971). This is still a steady state flow wherein toroidal vortices encircle the rotating shaft (Dumont et al., 2000c). Further increase in rotational velocities leads to unstable and wavy vortices. These flow regimes are very well studied for annular systems in the absence of blades (Dumont et al., 2002).
The first attempts to visualize flow patterns in an SSHE were carried out by Trommelen and Beek in the early 1970s (Trommelen and Beek, 1971). Streamlines were visualized by injecting colored liquid as a tracer. For laminar flow, the streamlines were analyzed for two blades in closed (no-slip) and open (annular) configurations. For Tar < Tar,Cr, they found that the streamlines were concentric circles in the open positions suggesting classic flow under shear. For no-slip condition, streamlines formed closed loops between the two blades suggesting pressure driven flow. No flow visualizations were performed for the conditions where Tar > Tar,Cr (Trommelen and Beek, 1971). In the past, transition of flow patterns in SSHEs have been analyzed for annular gaps without blades (Harrod, 1986; Trommelen and Beek, 1971). In addition, several studies concluded that the flow patterns with or without blades in the annulus were fundamentally similar (Abichandani et al., 1987; Harrod, 1986). Dumont and co-workers (Dumont et al., 2000b,c) studied the effects of blades on flow regimes using visual as well as electrochemical techniques. Based on visual presence of Taylor toroidal vortices, and diffusion limited electrochemical outputs, they concluded that the flow regimes were fully controlled by the rotating blades for both Newtonian and pseudoplastic fluids (Dumont, et al., 2002). The vortex flow patterns were renewed at each blade passage and the critical values of Tar for systems with blades were significantly higher than those without blades.
Figure 2a depicts a schematic for the vortex flow patterns for a Newtonian fluid in an annulus. Depending on the directionality of the vortical flow, maximum (SM) and minimum (Sm) shear stresses are observed. For a non-Newtonian CMC solution, the Taylor vortices are deformed suggesting a strong dependence of fluid properties on the flow pattern.
Stranzinger and co-workers extensively studied flow patterns, shear stress profiles, and scraper blade pressure effects in model SSHEs (Stranzinger et al., 2002; Stranzinger et al., 2001). In their investigations, the rotational velocities were kept below the requirements of Couette to Taylor vortical flow transition limits (Stranzinger et al., 2001). Shear-thinning and Newtonian fluids were analyzed by using numerical flow simulations, and experimental flow visualizations in a two-dimensional plane.
Three-dimensional effects like Taylor vortices were eliminated from their studies by using low rotational velocities. Similar to the results obtained by Dumont and co-workers (Dumont et al., 2000c), they found that the flow patterns around the scraper blades were controlled predominantly by their presence. Numerical simulations involving viscous heat dissipation also revealed higher shear stresses around the scraper blades (Stranzinger et al., 2001). By varying system parameters like rotor velocity, angle of scraper blades, and the scraper blade gap, the flow patterns could be influenced significantly (Dumont et al., 2000c; Stranzinger et al., 2001). For highly temperature sensitive processes, a combination of shorter blades with low rotor velocities was suggested for homogeneous temperature fields. Similarly, adjustment of the blade angle and the blade type were determined to be powerful design parameters for specific food microstructure requirements. For example, short scraper blade length was found to be more conducive for emulsifying flow processes where an efficient breaking of droplets is desired (Stranzinger et al., 2002). Similarly, for crystallization processes, shear structuring of crystals is desirable and hence, a blade angle perpendicular to the axis is more suitable (Stranzinger et al., 2002).
Dumont and co-workers extensively studied the effects of blades on shear rates in Couette and Taylor vortex regimes for Newtonian and various non-Newtonian fluids (Dumont et al., 2000a,b,c; Dumont et al., 2002). Shear rates were measured electrochemically in simple annular flow and in SSHE with two blades (Dumont et al., 2000b). They found that the shear rates measured in scraped systems were 10–100 times higher than those measured in an annulus without scrapers. For annular systems with Tar less than the critical value (Tar,Cr), the ratio of shear rate to rotational speed (Sr/Nr) was constant with increasing Tar. The transition from Couette flow to Taylor vortices was evident in an increase in the Tar as function of Sr/Nr beyond the Tar,Cr. However, in the presence of scraping, this transition was not evident from the shear rate data. For the system used by Dumont and co-workers, it appeared that the presence of blades dominated the shear rate profile (Dumont et al., 2000a,b,c; Dumont et al., 2002).
Investigations using starch suspensions as the product revealed that the presence of blades indeed subjected large stresses on the pumped product (Mabit et al., 2003). Mechanical swelling of starch was used to quantify the volume fraction of processed product undergoing high shear rates. In the absence of blades for flow regimes with Tar greater than Tar,Cr, shear rates achieved were below the threshold shear rates required to cause swelling in starch granules. In the presence of two blades perpendicular to the axis, however, an increasing volume fraction of starch was subjected to high shear as evidenced by swelling (Mabit et al., 2003).
Flow patterns in rotating geometries like annular spaces with or without blades, have also been investigated using nuclear magnetic resonance imaging (MRI) (Corbett et al., 1995). Using model Newtonian systems, they showed that the presence of blades in the SSHE removed the concentration inhomogeneities. The flow patterns in the Couette flow regime were similar to those found in previous investigations (Dumont et al., 2000c; Harrod, 1986; Trommelen and Beek, 1971).
Wang and co-workers studied the effect of shear-thinning behavior on the flow patterns in a single-bladed SSHE operating in the laminar regime (Wang et al., 1999; Wang et al., 2000). Studying the flow patterns for different power law fluids using nuclear magnetic resonance, they also found that the flow patterns in the presence of blades were different than simple annular flow patterns. The velocity profile for SSHE with one blade exhibited a parabolic profile with a maximum value at approximately one third the distance between the inner and the outer cylinder. For increasing shear-thinning nature of the fluids, the velocity maximum moved toward the outer cylinder. They also observed that for extremely shear thinning fluids, in their case suspended tomato puree, particle migration away from the high shear occurred resulting in serum separation. This also corresponded with no-slip condition at the outer tube wall (Wang et al., 1999; Wang et al., 2000).
Flow patterns in SSHEs are complicated because of simultaneous axial flow and rotation of blades. The presence of blades dominates the evolution of flow patterns and shear profiles. For fluids whose viscosity is dependent on shear stresses during the flow (for example shear-thinning), the presence of rotating blades deforms the flow patterns, making prediction of residence time distribution extremely difficult