In this numerical study, a new type of serpentine micromixer involving mixing units with a non-rectangular cross-section is investigated. Similar to other serpentine/spiral shaped micromixers, the design exploits the formation of transversal vortices (Dean flows) in pressure-driven systems, associated with the centrifugal forces experienced by the fluid as it is confined to move along curved geometries. In contrast with other previous designs, though, the use of non-rectangular cross-sections that change orientation between mixing units is exploited to control the center of rotation of the transversal flows formed. The associated extensional flows that thus develop between the mixing segments complement the existent rotational flows, leading to a more complex fluid motion. The fluid flow characteristics and associated mixing are determined numerically from computational solutions to Navier–Stokes equations and the concentration-diffusion equation. It is found that the performance of the investigated mixers exceeds that of simple serpentine channels with a more consistent behavior at low and high Reynolds numbers. An analysis of the mixing quality using an entropic mixing index indicates that maximum mixing can be achieved at Reynolds numbers as small as 20 in less than four serpentine mixing units.

Computational fluid dynamics modeling at Reynolds numbers ranging from 10 to 100 was used to characterize the performance of a new type of micromixer employing a serpentine channel with a grooved surface. The new topology exploits the overlap between the typical Dean flows present in curved channels due to the centrifugal forces experienced by the fluids, and the helical flows induced by slanted groove-ridge patterns with respect to the direction of the flow. The resulting flows are complex, with multiple vortices and saddle points, leading to enhanced mixing across the section of the channel. The optimization of the mixers with respect to the inner radius of curvature (R_in) of the serpentine channel identifies the designs in which the mixing index quality is both high (M > 0.95) and independent of the Reynolds number across all the values investigated.

A computational model has been developed to investigate how the magnetostatic interactions affect the hysteresis and magnetization curves for hexagonal arrays of magnetic nanowires. The magnetization coupling between nanowires arises from the stray fields produced by the other nanowires composing the array such that the field at each nanowire is the sum of the external field and the interaction field with the other nanowires. Using only two adjustable parameters: the interaction between nearest neighbors and the width of the Gaussian distribution in switching fields centered around the measured coercivity, simulations are compared with the experimentally measured hysteresis and magnetization curves for electrodeposited ${\mathrm{Co}}_{\text{0.45}}$ ${\mathrm{Fe}}_{\text{0.55}}$ alloy nanowires with diameters from 12 to 48 nm. Excellent agreement is found for all nanowire systems except for the largest diameter arrays where deviations from the Gaussian distribution of switching fields need to be considered.

We explore numerically the feasibility of enhancing the mixing capability of microchannels by employing the Weierstrass fractal function to generate a pattern of V-shaped ridges on the channel floor. Motivated by experimental limitations such as the finite resolution (~10 μm) associated with rapid prototyping through soft lithography techniques, we study the influence on the quality of mixing of having finite width ridges. The mixing capability of the designs studied is evaluated using an entropic measure and the designs are optimized with respect to: the distances between the ridges and the position range of their tip along the width of the channels. The results are evaluated with respect to the benchmarks established by the very successful staggered herring bone (SHB) design. We find that the use of a non periodic protocol to generate the geometry of the bottom surface of the microchannels can lead to consistently larger entropic mixing indices than in cyclic structures. Furthermore, since the optimization curves (mixing index vs. geometric parameters) are broader at the maximum for fractal microchannels than for their SHB counterparts, the microchannel designs using the Weierstrass fractal function are less sensitive to experimental uncertainties.

The fluid flow and mixing characteristics of planar serpentine microchannels consisting of repeating S-shaped units was characterized both computationally and experimentally for Reynolds numbers Re < 100. In particular, it was found that using non-rectangular cross-sections with changing orientation for the geometrical units greatly enhances their mixing performance relative to standard rectangular profile serpentine channels. The optimization of the cross-section enables efficient mixing of two-fluid components at Reynolds numbers as low as 20, making these designs practical for chemical reaction engineering and biological assays platforms. A computational analysis of the cross-sectional flow structures present in this type of channels indicates that their enhanced performance is correlated with the formation of secondary transversal vortexes. In the optimal designs, their presence increases the mass transfer in the microchannel by both stretching the interface between the fluid components to be mixed, as well as by promoting chaotic advection.

The magnetic interactions in large arrays of ordered magnetic nanowires with 12–48nm diameter and 55–95nm spacing were investigated using modified Henkel plots. The measurements for nanowire arrays ac demagnetized with the field applied parallel to the nanowire axis (the easy magnetization axis) indicate that the dominant interaction during the switching process is the magnetostatic coupling between the nanowires. Nevertheless, while the strength of the magnetostatic interactions increases with the magnetic moment associated with the nanowires, the increase is not linear with respect to the volume of the nanowires. Moreover, the dependence of the remanence curves on the field history suggests that even for magnetic nanowire systems with high geometric anisotropy, the magnetic pole structure of the nanowires can be complex. This conclusion is also supported by the field dependence of the initial magnetization curves.