The acoustically-driven microfluidic extensional rheometer: development, validation, and application to complex low-viscosity fluids

2017-03-03T06:06:49Z (GMT) by McDonnell, Amarin George
The emergence of rheometric techniques for extensional flows is relatively recent (Bazilevsky, 1990; McKinley, 2002) when compared to methods for shear flows (Dontula, 2005). Two prominent extensional rheometry techniques involve creating liquid filaments that thin under the action of surface tension. Stress balances (Tirtaatmadja, 1993; Szabo, 1997; McKinley, 2002; McKinley, 2000) are used to determine the extensional viscosity from measurements made on such filaments. In the filament stretching extensional rheometer (FiSER – Tirtaatmadja 1993) a fluid sample is placed between two end-plates that are mechanically drawn apart at a controlled exponential rate. Another prominent technique, capillary break-up extensional rheometry (CaBER – McKinley, 2000; Rodd, 2004), also employs mechanical end-plates which are rapidly moved to a fixed separation to study the dynamics of liquid bridges. These two techniques have been successful in assessing the extensional properties of a wide variety of complex fluids from polymer solutions to suspensions (Tirtaatmadja, 1993; McKinley, 2002; Ooi, 2004). Although these two uni-axial extensional flow techniques have gained acceptance they are limited in their utility when analysing low-viscosity fluids. The stresses that occur in these flows can be non-dimensionalised so that the importance of different effects can be established (Rodd 2004). These experiments tend to use larger sample sizes, which can lead to an adverse asymmetry in the filament in which gravity causes the filament to sag. The mechanical operation of these devices can induce vibrations that propagate throughout the filament, which are not accounted for by stress balance analysis, and are not damped sufficiently in low-viscosity fluids. Such inertial effects are exacerbated by the large samples sizes. These issues make repeatable readings impractical in low-viscosity fluids (Rodd, 2004). Recent modifications of the CaBER concept permit improved access to low-viscosity fluids (Vadillo, 2010; Vadillo, 2012; Campo, 2010; Nelson, 2011; Keshavarz, 2015; Arratia, 2008). Nevertheless, it is still a challenge to obtain viscosities of complex aqueous fluids of viscosities of the order of 1 mPa.s. This thesis proposes and demonstrates a new microfluidic technique for extensional rheometry that harnesses the unique capabilities afforded by surface acoustic waves (SAW) to generate liquid filaments that thin under surface tension effects. This approach has three key attributes that are advantageous for the analysis of thin and complex fluids in extensional flow: - SAWs are able to generate stable fluid filaments from low-viscosity fluids. - Small sample volume sizes mean that effects due to gravity and inertia scale down favourably. Critically, this also permits the analysis of materials that are difficult to obtain in large volumes. - SAWs have the ability to manipulate fluids whilst not damaging delicate fluid constituents. This technique is used to analyse many novel fluids for a variety of purposes: validating theoretical predictions, optimising promising industrial fluids, and providing insights into unusual phenomena. These studies serve to validate the acoustically-driven capillary break-up approach as a valuable rheometric technique. The thesis is presented in the "thesis including published works" format wherein publications are combined with explanatory notes and additional unpublished material The dissertation begins by providing a background on rheology, and explains the need for extensional experiments to extract the extensional properties of fluids. It also reviews the limitations of the aforementioned conventional extensional techniques, specifically the problems they encounter when analysing low-viscosity fluids. Additionally, it defines the two fundamental challenges to be overcome in a filament-based extensional rheometer: the creation of a filament to reliably measure the neck radius as a function of time, and converting this data into a measurement of rheological properties. The novel ways in which SAWs manipulate fluids is then discussed, and their ability to generate fluid filaments is proposed as a means for extensional rheometry of low-viscosity fluids. The next chapter discusses the experimental approach that addresses the first of the fundamental challenges mentioned in the chapter above. It examines the obstacles and solutions involved with the development of the acoustically-driven capillary break-up device and those pertaining to the analysis of the low-viscosity fluid filament break-up data that it produces. Experimental challenges include the fabrication, operation, and testing of both the SAW device and the experimental rig which harnesses the device. Other challenges are the high spatial and temporal data resolution needed to distinguish differences between microfluidic break-up events of different but similar thin fluids. Despite these difficulties in development, the experimental system is shown to be suitable for rheological analysis and possesses considerable advantages, including the use of small test volumes, the stabilisation of thin filaments by SAWs, and the ability to harmlessly manipulate delicate particle suspensions and macromolecular solutions. The following chapter is in the form of a publication (Bhattacharjee, 2011) that demonstrates the validity of the acoustically-driven capillary break-up technique for rheological measurement. It also defines the conditions under which SAWs can be used to create liquid filaments, complementing the previous chapter. However, the main focus is the comparison of our experimental results with data from accepted rheological experiments, and the quantification of our data using standard rheological analysis. Good agreement is found from necking data of low-viscosity Newtonian and non-Newtonian fluids. Moreover, it is demonstrated that stress-balance analysis can be used here to convert the filament decay data from a viscoelastic strain-hardening fluid directly to stress and viscosity measurements. Finally, the technique is utilised in observing the effects of extensional flow on a small volume of low-viscosity solution containing a dissolved protein, for which measurements showed an interesting multi-stage filament break-up process. The subsequent chapter discusses the generic difficulties of extracting extensional viscosity using stress-balance analysis from the capillary break-up of low-viscosity fluids that do not exhibit strain-hardening; it then develops a new approach for such fluids. Analysis using a simple mid-filament stress balance, as seen in the previous chapter, is appropriate when thread-like filaments are formed; this occurs in the break-up of viscoelastic solutions, where long-lived filaments develop towards final break-up. However, many other fluids, particularly low-viscosity solutions, do not form near-cylindrical filaments. This leads to stress-balance analysis being complicated by the dynamic contributions of axial filament curvature, the difficulties of which are investigated here. Thus, extensive non-dimensionalised break-up data of Newtonian fluids is used to develop a calibration method that allows the extraction of extensional viscosity for unknown fluids without resorting to the prohibitively complicated approach of using full filament analysis to determine extensional viscosity. The next chapter is a publication (McDonnell, 2015) where the calibration method of extracting extensional viscosities is used with data from the acoustically-driven capillary break-up technique to pursue the validation of previously untested theoretical predictions of "active matter" bulk properties in extensional flow. Active matter theory (Hatwalne, 2004; Ramaswamy, 2010; Marchetti, 2013) describes suspensions composed of individual particles that are self-propelled, where their net average alignment contributes to the overall stresses and viscosity of the suspension. This theory represents active bodies as axisymmetric particles that exert a net thrust along their primary axis, producing hydrodynamic dipoles that drive the surrounding fluid along their lengths. Active particles can be placed into two groups: "pushers" drive a tensile flow along their principal axis (Drescher, 2011), while conversely "pullers" generate a contractile flow (Hatwalne, 2004); the positive hydrodynamic dipoles created by pusher particles are predicted to lower suspension viscosity below that of an equivalent passive particles suspension, while the negative hydrodynamic dipoles of pullers will increase viscosity. Predictions have been validated by experimental findings for pushers (Sokolov, 2009; Gachelin, 2013; Karmakar 2014) and pullers (Rafai, 2010) in shear flow, but not in extensional flow. Such systems are exemplified by living materials, like suspensions of motile microbes or the cytoskeletal polymers in ATP-powered white blood cells. However, such delicate biological suspensions are difficult to prepare in high volumes, and typically require aqueous media to suspend them and enable their motility, aspects that pose problems for alternative extensional techniques but are accommodated by the acoustically-driven capillary break-up technique. We measure the extensional viscosities of pushers using bacterial and mouse sperm suspensions, and pullers using algal suspensions. We extend a model previously proposed for dilute suspensions of active particles to obtain predictions at moderate and high concentrations. The experimental data is shown to be in good agreement with these predictions. The comparison identifies some key parameters that are geometry dependent that can strongly influence the extensional viscosity of these materials. (...)