When a highly elastic drop of a polymer solution hits a superhydrophobic surface at a high speed, a growing tail-like filament emerges vertically from the impact spot as the contact line recedes. Notably, the ligament transitions into a balloon-like shape before detaching completely from the surface (Balloon regime). The ligament formation is attributed to liquid impalement upon impact into the surface protrusion spacing, and elastic forces due to polymers prevent ligament breakup. The detachment of the ligament happens when polymeric stresses balance or overcome the adhesion at the surface. This study shows that tuning droplet rheology and surface roughess enables droplets to rebound completely and without splashing at high impact speeds. 

Elasto-inertial turbulence (EIT) offers a mechanism to tune fundamental flow processes using elastic instabilities by the addition of polymer to the fluid. Here, we extend this emerging paradigm to elastoviscoplastic (EVP) materials, where the interplay of elasticity and a yield stress dictates the flow dynamics. Through direct numerical simulations of channel flow, we demonstrate that a yield stress does not suppress EIT but rather reorganizes it, with the flow maintaining an EIT-like characteristic even when most of the domain is nominally unyielded. Plasticity promotes the formation of solid-like plugs and patches, weakens near-wall streaks, and non-monotonically modifies drag, while elastic stresses continue to drive fluctuations in yielded shear layers. Detailed analysis of solid-fraction statistics, first-normal-stress fields, turbulent kinetic energy spectra and budgets jointly reveal a robust elastic turbulent core where EVP stresses transition from net sinks to net sources of turbulent kinetic energy as elasticity of the material increases. The turbulent kinetic energy spectra retain inertial and elastic scaling ranges, and flow topology collapses towards quasi-two-dimensional, sheet-like structures for large elasticity of the EVP material. Together the present results identify EIT in EVP fluids as an elastic turbulent state  modified by plasticity, bridging the gap between viscoelastic EIT and high–Reynolds-number EVP turbulence and providing mechanistic insight directly relevant to controlling drag, mixing and transport in yield-stress fluids.

Non-Newtonian fluids such as viscoelastic and elastoviscoplastic (EVP) materials are ubiquitous in industrial and environmental flows and exhibit complex rheological behavior that alters flow topology across multiple scales in comparison to Newtonian liquids. Their dynamics are governed by additional elastic and plastic stresses, which crucially affect the regimes of interfacial processes and turbulent flows. This thesis investigates how elasticity and yield stress jointly influence multiphase and canonical wall-bounded flows, and how data-driven methods can be used to infer experimentally inaccessible stress fields. The study is based on direct numerical simulations (DNSs) of viscoelastic and EVP flows, along with baseline convolutional neural networks for non-intrusive sensing in turbulent configurations.

We first investigate the effects of complex rheology on the dynamics of an interface in an EVP material when a bubble bursts on the surface, where the driving mechanism is the capillary stress. The numerical simulation of bubble bursting uncovers how the elastic stresses and their relaxation interacts with the yield stress. The parameter space in the study is spanned by the Deborah number and plasto-capillary number. Four distinct regimes are identified, ranging from Newtonian-like jetting to yield-dominated states with limited jet formation.

Then, we examine what happens when a viscoelastic droplet of polyacrylamide solution impacts a super-hydrophobic surface at high speed. The experimental investigation reveals that, in contrast to previous studies, viscoelastic droplets may experience full rebound while Newtonian droplets do not. This happens through a balloon regime in which the impacting droplet first impales the surface and the rebounding droplet forms a thin ligament which later rebounds together with the drop, like a rubber band. The regime arises from the combined effect of substrate characteristics, liquid impalement and strong polymeric stresses that prevent ligament breakup. Using DNS, we show that such a regime is reproduced numerically, provided sufficient modeling of the contact line behavior. Using a simplified theoretical framework we propose a necessary condition for the complete droplet rebound. In the splashing regime of Newtonian fluids, viscoelastic drop impact is shown to exhibit elongated fingers. Increasing polymer concentration enhances viscous damping, reducing the number of fingers and eventually suppressing fingering instabilities while preserving complete rebound. The onset and evolution of fingering are a result of the competition between inertia--capillary--viscous terms, and a theoretical framework is shown to predict the temporal growth of the characteristic ligament length across rheological conditions.

We then turn to a chaotic flow regime, and study elastoviscoplastic channel flow at low Reynolds numbers (Re=5-500), where viscoelastic fluids display elasto-inertial turbulence ("early turbulence"). We demonstrate that the presence of yield stress does not relaminarize the flow unlike in inertial turbulence at high Reynolds numbers. Further, we show that the EVP stresses transition from net sinks to net sources of turbulent kinetic energy as elasticity of the material is increased.

Since the turbulent nature of non-Newtonian fluid flows require description of both the flow and extra stresses, where the latter is not accessible in experiments, we develop a methodology to retrieve the velocity and elastic stress fields in viscoelastic turbulent channel flows from wall measurements. We show that the introduced baseline methodology can learn highly non-linear mapping between wall information and near-wall flow and stress fields but also is limited by the accurate retrieval of small scale structures. Towards enhancing the performance of the non-intrusive sensing methodology in an experimental setting, we introduce spectrally informed loss functions which are tested in a Newtonian fluid flow configuration as a part of this thesis work and will be extended to non-Newtonian configurations. Taken together, the findings in the present thesis provide an idea of the behavior of complex fluids in different configurations, while demonstrating the potential of data-driven techniques to estimate inaccessible stress fields in non-Newtonian turbulence.

Icke-newtonska vätskor, såsom viskoelastiska och elasto-viskoplastiska (EVP) material, är vanliga inom olika industrier och i naturen. De har komplexa flödesbeteenden som — jämfört med newtonska — vätskor — kan förändra flödet över flera skalor. Dessa vätskors dynamik styrs av ytterligare elastiska och plastiska spänningar, vilket gör att flödet kan ändra karaktär (regim) både i gränsytfenomen och i turbulenta flöden. Denna avhandling undersöker hur elasticitet och flytspänning (yield stress) tillsammans påverkar multifasflöden och kanoniska väggbundna flöden, samt hur datadrivna metoder kan användas för att uppskatta spänningsfält som inte är experimentellt åtkomliga. Studien bygger på direkta numeriska simuleringar (DNS) av viskoelastiska och EVP-flöden, i kombination med grundläggande konvolutionella neurala nätverk (CNN) för icke-intrusiv avkänning i turbulenta konfigurationer.

Vi studerar först hur komplex reologi påverkar dynamiken av den fria ytan av ett EVP-material när en luftbubbla brister på ytan i ett EVP-medium, där den drivande kraften kommer från ytspänningen. DNS av bubbelbristning avslöjar hur elastiska spänningar och deras relaxation interagerar med flytspänningen i materialet-. Parameterutrymmet i studien spänns upp av Deborah-talet och det plasto-kapillära talet. Fyra distinkta regimer identifieras, från newtonsk-liknande jetbildning till flytspänningsdominerade tillstånd med begränsad jetutveckling.

Därefter analyseras vad som händer när en viskoelastisk droppe av polyakrylamidlösning träffar en superhydrofob yta med hög hastighet. Experimenten visar att viskoelastiska droppar — i kontrast till tidigare studier — kan återstudsa fullständigt även när newtonska droppar inte gör det. Detta sker via en ballongregim, där den återstudsande droppen formar ett tunt ligament som först tränger genom ytan, och sedan lossnar tillsammans med droppen likt ett gummiband. Regimen uppstår genom samspelet mellan ytegenskaper, vätskans genomträngning och starka elastiska spänningar som motverkar ligamentets avbrott. Med DNS visar vi att denna regim kan reproduceras numeriskt, förutsatt att kontaktlinjens dynamik modelleras tillräckligt väl. Med ett förenklat teoretiskt ramverk föreslår vi dessutom ett nödvändigt villkor för fullständig droppåterstuds.

I stänkregimen (splashing) för newtonska vätskor uppvisar viskoelastiskt droppnedslag förlängda fingrar. Ökad polymerkoncentration förstärker den viskösa dämpningen, vilket minskar antalet fingrar och till slut undertrycker fingringsinstabiliteter, samtidigt som fullständig återstuds bibehålls. Fingringens uppkomst och utveckling bestäms av konkurrensen mellan tröghets-, kapillär- och viskösa bidrag, och ett teoretiskt ramverk visas kunna förutsäga den tidsmässiga tillväxten av den karakteristiska ligamentlängden över olika reologiska förhållanden.

Vi övergår till ett kaotiskt flödesregim och studerar elasto-viskoplastisk kanalströmning vid låga Reynoldstal (Re=5-500), där viskoelastiska vätskor uppvisar elasto-inertiell turbulens ("early turbulence"). Vi visar att närvaron av flytspänning inte relaminariserar flödet, till skillnad från vad som kan observeras i inertiell turbulens i höga Reynoldstal. Vi visar vidare att EVP-spänningarna övergår från att vara nettosänkor till att bli nettokällor för turbulent kinetisk energi när materialets elasticitet ökar.

Eftersom turbulent icke-newtonsk strömning kräver en beskrivning av både hastighetsfältet och de extra spänningarna — där de senare i praktiken inte är experimentellt åtkomliga-utvecklar vi en metodik för att rekonstruera hastighets- och elastiska spänningsfält i viskoelastiska turbulenta kanalflöden utifrån väggmätningar. Vi visar att den föreslagna baslinjemetodiken kan lära sig en starkt icke-linjär avbildning mellan vägginformation och väggnära flödes- och spänningsfält, men att den samtidigt begränsas av förmågan att korrekt återge småskaliga strukturer. För att förbättra prestandan hos den icke-intrusiva avkänningsmetodiken i en experimentell kontext introducerar vi spektralt informerade förlustfunktioner, som testas i en newtonsk flödeskonfiguration inom ramen för avhandlingsarbetet och därefter kommer att utvidgas till icke-newtonska konfigurationer. Sammantaget ger resultaten i avhandlingen en samlad bild av hur flytspänningsvätskor beter sig i olika konfigurationer, samtidigt som de visar potentialen hos maskininlärningsbaserade, datadrivna metoder för att uppskatta annars otillgängliga spänningsfält i icke-newtonsk turbulens.

Controlling the motion of non-Newtonian drops on surfaces is crucial for applications ranging from inkjet printing to biomedical devices and food processing. While the macroscopic behavior of viscoelastic drops sliding on tilted hydrophobic surfaces has been characterized—showing reduced velocities and elongation compared to Newtonian fluids—the underlying microscopic mechanisms remain poorly understood. To address this gap, we developed a high-speed, high-resolution reflection microscope that enables direct visualization of the contact line of sliding drops. We used water-soluble polyelectrolyte solutions based on polyacrylamide and allowed drops to slide on hydrophobic substrates composed of Teflon amorphous fluoropolymer and polydimethylsiloxane-coated glass slides. The substrate tilting angle was varied between 20° and 45°. We reveal how viscoelasticity influences the dynamics of the receding contact line and drop motion. Our experiments demonstrate that viscoelasticity can destabilize the receding contact line, triggering filament formation. This instability previously observed in the coating of thin viscoelastic films is reported here for sliding drops, for the first time on smooth surfaces. We further highlight the critical role of polymer charge in this process: while cationic and nonionic polymers promote filament formation, anionic polymers do not, a difference we attribute to the distinct wetting properties of the solutions. In conclusion, we clarify the interplay between rheology, surface interactions, and drop dynamics.

Accurate estimation of near-wall turbulence from limited surface measurements is critical for various practical and experimental applications; however, it remains challenging due to the complex and multiscale characteristics inherent to turbulent flows. Recent developments in data-driven methods have improved performance over traditional linear approaches, but are often limited by loss functions that primarily penalize point-wise errors. In this study, we consider a composite, spectrally informed loss function that augments the conventional mean-squared error with terms that explicitly promote statistical consistency of fluctuation levels and spectral energy distributions. This loss function is evaluated using a baseline convolutional network model, applied to direct numerical simulation datasets of turbulent open-channel flow at friction Reynolds numbers, Re_τ =180 and 550. We show that considered loss function substantially reduces reconstruction errors of near-wall fluctuation velocity fields obtained from the network model using wall-shear stress and wall-pressure data as inputs. The proposed method recovers up to a threefold improvement in reconstruction accuracy relative to baseline employing mean-squared loss, and retains energy content at small-scale structures. We further show that reconstruction accuracy is only mildly affected when the input wall signals are contaminated with 5-100% Gaussian noise, and retains energy content in small flow structures when predicting from coarse wall inputs, indicating that the trained models generalize robustly to noisy, experimentally relevant conditions. Our results demonstrate that properly designed loss functions can improve reconstruction accuracy, highlighting the importance of training strategy in achieving efficient, high-performance neural-network models for practical non-intrusive sensing applications. The present study also provides a foundation for future extensions with advanced architectures, such as generative or diffusion-based models to further improve reconstruction of turbulent flow fields in complex settings.

 When water droplets impact solid surfaces at high velocity, they often develop radial protrusions—known as fingering instabilities—that subsequently break up during spreading and retraction, a process termed splashing. Here, we investigate the fingering dynamics of shear‑thinning viscoelastic droplets impacting superhydrophobic surfaces. At low polymer concentrations, liquid elasticity sustains the emergence of elongated fingers, while simultaneously stabilizing them against breakup, thereby suppressing splashing. In contrast, increasing polymer concentration enhances viscous damping, reducing the number of fingers and ultimately suppressing the fingering instability. Our results indicate that the onset of fingering is governed by the interplay of inertia, surface tension, and viscous stresses, while the number of fingers scales robustly with the Weber number. This highlights the dominance of inertia-capillary dynamics in our range of Weber number once the instability is triggered. Remarkably, all impact outcomes resulted in complete rebound, in contrast to previous observation for viscoelastic droplets. Finally, we employ a theoretical framework to predict the temporal evolution of the mean ligament length across polymer concentrations, providing quantitative insight into how elasticity modifies drop retraction dynamics.

When water droplets impact solid surfaces at high velocity, they often develop radial protrusions, known as fingering instabilities, that subsequently break up during spreading and retraction, a process termed splashing. Here, we investigate the fingering dynamics of shear-thinning viscoelastic droplets impacting superhydrophobic surfaces. At low polymer concentrations, liquid elasticity promotes the emergence of elongated fingers while simultaneously stabilizing them against breakup, thereby suppressing splashing. In contrast, an increasing polymer concentration enhances viscous damping, reducing the number of fingers and ultimately suppressing the fingering instability. Our results indicate that the onset of fingering is governed by the interplay of inertia, surface tension, and viscous stresses, while the number of fingers scales robustly with the Weber number. This highlights the dominance of inertia–capillary dynamics in our range of Weber numbers once the instability is triggered. Remarkably, all impact outcomes resulted in complete rebound, in contrast to a previous observation for viscoelastic droplets. Finally, we employ a theoretical framework to predict the temporal evolution of the mean ligament length across polymer concentrations, providing quantitative insight into how elasticity modifies drop retraction dynamics.

Fully-convolutional neural networks (FCN) were proven to be effective for predicting the instantaneous state of a fully-developed turbulent flow at different wall-normal locations using quantities measured at the wall. In Guastoni et al. (J Fluid Mech 928:A27, 2021. https://doi.org/10.1017/jfm.2021.812), we focused on wall-shear-stress distributions as input, which are difficult to measure in experiments. In order to overcome this limitation, we introduce a model that can take as input the heat-flux field at the wall from a passive scalar. Four different Prandtl numbers Pr=ν/α=(1,2,4,6) are considered (where ν is the kinematic viscosity and α is the thermal diffusivity of the scalar quantity). A turbulent boundary layer is simulated since accurate heat-flux measurements can be performed in experimental settings: first we train the network on aptly-modified DNS data and then we fine-tune it on the experimental data. Finally, we test our network on experimental data sampled in a water tunnel. These predictions represent the first application of transfer learning on experimental data of neural networks trained on simulations. This paves the way for the implementation of a non-intrusive sensing approach for the flow in practical applications.

Neural network models have been employed to predict the instantaneous flow close to the wall in a viscoelastic turbulent channel flow. Numerical simulation data at the wall are used to predict the instantaneous velocity fluctuations and polymeric-stress fluctuations at three different wall-normal positions in the buffer region. Such an ability of non-intrusive predictions has not been previously investigated in non-Newtonian turbulence. Our comparative analysis with reference simulation data shows that velocity fluctuations are predicted reasonably well from wall measurements in viscoelastic turbulence. The network models exhibit relatively improved accuracy in predicting quantities of interest during the hibernation intervals, facilitating a deeper understanding of the underlying physics during low-drag events. This method could be used in flow control or when only wall information is available from experiments (for example, in opaque fluids). More importantly, only velocity and pressure information can be measured experimentally, while polymeric elongation and orientation cannot be directly measured despite their importance for turbulent dynamics. We therefore study the possibility to reconstruct the polymeric-stress fields from velocity or pressure measurements in viscoelastic turbulent flows. The neural network models demonstrate a reasonably good accuracy in predicting polymeric shear stress and the trace of the polymeric stress at a given wall-normal location. The results are promising, but also underline that a lack of small scales in the input velocity fields can alter the rate of energy transfer from flow to polymers, affecting the prediction of the polymeric-stress fluctuations.

A gas bubble sitting at a liquid-gas interface can burst following the rupture of the thin liquid film separating it from the ambient, owing to the large surface energy of the resultant cavity. This bursting bubble forms capillary waves, a Worthington jet and subsequent droplets for a Newtonian liquid medium. However, rheological properties of the liquid medium like elastoviscoplasticity can greatly affect these dynamics. Using direct numerical simulations, this study exemplifies how the complex interplay between elasticity (in terms of elastic stress relaxation) and yield stress influences the transient interfacial phenomenon of bursting bubbles. We investigate how bursting dynamics depends on capillary, elastic and yield stresses by exploring the parameter space of the Deborah number ${{\textit {De}}}$ (dimensionless relaxation time of elastic stresses) and the plastocapillary number $\mathcal {J}$ (dimensionless yield-stress of the medium), delineating four distinct characteristic behaviours. Overall, we observe a non-monotonic effect of elastic stress relaxation on the jet development while plasticity of the elastoviscoplastic (EVP) medium is shown to affect primarily the jet evolution only at faster relaxation times (low ${{\textit {De}}}$). The role of elastic stresses on jet development is elucidated with the support of energy budgets identifying different modes of energy transfer within the EVP medium. The effects of elasticity on the initial progression of capillary waves and droplet formation are also studied. In passing, we study the effects of solvent-polymer viscosity ratio on bursting dynamics and show that polymer viscosity can increase the jet thickness apart from reducing the maximum height of the jet.