Error message

Monday, 20 January 2020
Time Speaker Title Resources
09:20 to 09:30 Organizers Introduction
09:30 to 09:45 Binod Sreenivasan (IISc., India) Large thermal anomalies in the lower mantleand implications for the geomagnetic field

A simplified two-fold pattern of convection in the Earth’s core is often used toexplain the non-axisymmetric magnetic flux concentrations in the present daygeomagnetic field. For large thermal anomalies in the lower-mantle, however,a substantial east-west dichotomy in core convection may be expected. In thisstudy, the effect of a large lateral variation at the outer boundary is examinedin (a) cylindrical annulus experiments that achieve approximate geostrophyof the convection and (b) a rapidly rotating spherical shell dynamo model.The pattern of convection in the dynamo simulations closely follows thatin the annulus experiments, which suggests that the lateral buoyancy at theequator essentially determines the structure of core convection. In particular,an isolated downwelling forms only beneath the Siberian region in stronglydriven convection. This can explain the relative instability of the magneticflux patches observed in the Western hemisphere.

09:45 to 10:00 Robin Joseph (IISc., India) Growth of disturbances in a pre transitional boundary layer downstream of distributed surface roughness.
10:00 to 10:15 Samson Annapu Reddy (IISc., India) Numerical Investigation on Transition Mechanisms in a Laminar Separation Bubble with Increasing Reynolds Number
10:15 to 10:30 Ritabrata Thakur (ICTS-TIFR, Bengaluru, India) Laminar shear flows often transition to turbulence at Reynolds numberssmaller than that of their first linear instability

Laminar shear flows often transition to turbulence at Reynolds numberssmaller than that of their first linear instability. The underlying mechanismcan be transient algebraic growth, either linear or nonlinear in nature. Weobtain the maximum perturbation energy growth in a three-dimensional heatedplane channel. A nonlinear technique of direct-adjoint-looping is employed tonumerically achieve this. With this technique, we also obtain the initial velocityand temperature perturbation structure that leads to this growth. The energygrowth associated with this optimal can be large enough to push the flow toturbulence. We study the effect of varying stratification strengths (gradientRichardson numbers), Reynolds numbers, and target times on the structureof the optimal. We show similarity or the lack thereof between the optimalperturbations for small (linear optimal) and large (non-linear optimal) initialenergies.

10:30 to 10:45 Udita Udya Ghosh (IISc., India) Slippery Rod-climbing effect?

Certain non-Newtonian fluids like hen’s egg, condensed milk, polymer melts when subjected to a partially immersed rotating rod exhibit ‘rod-climbing’. This spectacular phenomenon known as the Weissenberg effect has been so far attributed to the dominance of the normal stress differences over the rotation generated centrifugal forces. In this study, we delve into the question of whether these rod-climbing fluids continue their ascent on the rod in presence of a slippery interfacial layer (rod-fluid interface)? To execute this, immersed portion of the rod is covered with an oil layer (silicone and castor oils). Polyethylene oxide suspensions in deionized water are used as a model polymeric fluid in this study. In conjunction with this, the strength of first normal stress difference is also modulated to investigate if it affects the capacity of the polymer in sustaining the rod-climbing effect. This variation is brought about altering the polymer concentration in the aqueous suspension. Surprisingly, three states of positive Weissenberg (fluid-meniscus rise or classical rod climbing), negative Weissenberg (fluid-meniscus dip) and absence of Weissenberg effect are observed. The primary centrifugal force induced shear deformation of the meniscus height is evaluated to characterize these states. A consolidated phase diagram based on the experimental results is proposed demarcating these states as a function of the polymer concentration, stirrer rotational speed (or applied shear) and type of oil layer. However, the rod-climbing effect is also known to be accompanied by the secondary flows. Dye visualization experiments are performed to qualitatively map the secondary flows that show a characteristic flow direction depending on the state (positive/negative) exhibited by the fluid.

10:45 to 11:00 Y Chandukrishna (CSIR-NCL, India) Vortical structures arising out of mutual interaction between wing, tip mounted propellersand multiple propellers placed along the span of the wing

For the past few years, the aerospace industry has been talking about eco-friendly all-electric or hybridelectric aircraft. In the case of all-electric aircraft, power available is a constraint, so, a lot of designmodifications, multi-disciplinary optimizations have to be performed at the design phase. Some designswhich are considered by the industries are wing tip mounted propellers, multiple propellers along the span,V-tail propellers etc. In the present study, wing tip mounted propellers along with the multiple propellersalong the span of the wing are considered to examine the mutual interactions. In literature, a few people havestudied about the benefits of wing tip mounted propeller. Need has arrived now to restart the research againon these kinds of configurations.

In the first phase of study, propellers are approximated as actuator disks. Different configurations are beingtried to examine the benefits of wing tip mounted propellers when multiple propellers are used along thespan. CFD simulations are being performed on various configurations: isolated wing, propellers withoutwing, wing with one propeller at midspan, wing with one propeller at one third of the span from the root,wing with one propeller at the tip, wing with one propeller at the tip and one propeller at one third of thespan of the wing from root, wing with one propeller at the tip and two propellers at one third of the span andtwo third of the span from the root. RANS simulations were performed with symmetric wall boundarycondition for all the above cases using the open-source SU2 code. The phenomenon of vortex cancellation orweakening the strength of trailing vortex has to be studied when the tip vortices from the wing tip mountedpropeller are superimposed on the trailing vortices of the wing. In the second phase of study, plan is to use anactual propeller model to simulate the flow, by making use of sliding mesh interface technique in SU2 andstudy the propeller – propeller interaction along with their effects on thrust and power. Results are expectedto show a considerable drag reduction when wing tip mounted propellers are used and its effect on lift has tobe examined further.

11:00 to 11:30 Break Tea/Coffee Break and Posters
11:30 to 11:45 Arnab Mukherjee (IISc., India) The role of the centrebody wake on the precessing vortex core dynamics of a swirl nozzle

The precessing vortex core (PVC) phenomenon in swirling jets is a helical instability in the flow driven by the coherentprecession of the vortex breakdown bubble (VBB) around the flow axis, resulting in the helical rollup of the shear layer.This instability is driven mainly by flow processes in the region upstream of the VBB. Centrebodies, commonly employed incombustor nozzles create recirculation zones in their wake (CWRZ) that can interfere with VBB precession and hence suppressthe PVC. We study this phenomenon in a swirl nozzle with a centerbody whose end face is flush with the nozzle exit plane, usinglarge eddy simulations (LES) and linear hydrodynamic stability analysis for a nominal Reynolds numbers Re=48,767and82,751. For one of the Re=82,751cases the centrebody end face diameter halved. This case shows the presence of coherentVBB precession in the flow. For the other two cases, linear stability analysis shows the presence of a stable VBB precessionmode whose stability characteristics are sensitive to the time averaged flow structure in the region where the VBB and CWRZmerge. This shows that the reduction in the CWRZ length, due to halving the centrebody end face diameter, causes the VBBprecession mode to become unstable. The oscillation frequency and the spatial distribution of oscillation amplitude predictedby linear stability analysis for this case, agree very closely with equivalent results determined from the LES using spectralproper orthogonal decomposition (SPOD). These results suggest that in general, spatial separation between the CWRZ andVBB can result in the onset of VBB precession and the emergence of PVC oscillations in swirl flows

11:45 to 12:00 Taye Melaku Taddesse (IISc., India) Numerical Study of Subsonic Twin Round Jets Using LES Based onExplicit Filtering

Multiple jets are found in several applications such as fuel injection, smoke stacks, jets enginesof aerospace vehicles, etc. Several researchers have experimentally studied the flow fielddevelopment of twin round jets. Okamoto et al. [1] carried out one of the earliest experimentalinvestigation of the interaction of parallel, twin turbulent round jets. The mean velocity and pressureprofiles were reported, but, turbulence quantities were not. They observed the shift in position ofmaximum velocity from the individual jet axes to the plane of symmetry of the jets. Rathakrishann,et al [2] also explored experimentally the flow field development of jets issuing from two circularnozzles set at a common end wall, with different nozzle separation distances. Their experimentshowed that the nozzle separation distance has a considerable influence on the development of thevelocity profiles up to downstream distances of 30d. Even a pair of round jets can have adistinctive development because, after the inevitable merging, there can be axis switching. LES ofparallel, twin jets reveal the structure and mechanisms of the flow-field. The LES is by an explicitfiltering method [3]. The twin round jets are at a Reynolds number Re of 230000, based ondiameter and mean velocity at exit. Distance between jet centers was 5 diameters. Closequantitative agreement with experiment was found on the development of mean profiles andspreading. Velocity fluctuations between the jets are weaker than those on the outer boundaries.Axis switching was observed.

12:00 to 12:15 Navneet Kumar (IISc., India) Revisitingcapillary rise experiments –Canevaporation limit it?
12:15 to 12:30 Suyog Vilasrao Mahulkar (IISc., India) Study of clapping propulsion by computational approach

In aquatic habitat, many species use pulse jet propulsion to obtain the thrust force. Thiskind of propulsion can be seen in a jellyfish where the contraction of the subumbrellar cavityresults in the production of thrust producing jet. The aim of the present study is to understandthe fluid dynamics of an impulsive force generation in the aquatic medium. The simplifiedexperimental model has been constructed using two thin flat plates interconnected with atorsion spring. The spring has held in the tension initially such that the angle between twoplates would be 60 degrees. Due to the sudden release of tension, both plates perform clappingaction in which the inter-plate angle starts reducing. During the clapping action of plates,fluid present in the inter-plate cavity will be ejected out and the body experiences an impulsivethrust force. Experiments have been conducted in the quiescent water medium.

The computational study of clapping propulsion has been performed using ‘ANSYS19’.The experimentally measured clapping body kinematics used as input to the computation. Thechallenges involved in the computational modeling of clapping action such as large displacementand contraction of the interplate cavity have been carefully handled using the ‘Overset Mesh’technique. The complicated 3D flow field has been studied computationally, which can bedifficult to analyze experimentally. The future study includes the investigation of the roleof added mass in such unsteady fluid dynamics. Based on the results of this computationalstudy, the wake of the clapping body, sufficiently downstream, can be approximately modeledas an ‘axisymmetric vortex’ ring, which is in the agreement with experiments. The involvedmomentum and energy balance in clapping propulsion may be studied using an ‘axisymmetricvortex ring model’.

12:30 to 12:45 Avinash Nayak (Academy of Scientific and Innovative Research, Delhi, India) Linear Stability Analysis of Transient Flows in a Duct
12:45 to 13:00 Chandan Kumar (Academy of Scientific and Innovative Research, Delhi, India) Asymmetric Nose-Fairing for Wing-Body Junction

An experimental investigation of the effect of elimination of the horseshoe vortex at a simplified wing-body junction, using an asymmetric nose-fairing is carried out. The wing used has a NACA0018 airfoil as cross-section and is mounted on a flat plate. Measurements are carried out at a low freestream velocity of 31m/s. A simple procedure is devised to design an asymmetric nose-fairing that could prevent saddle point separation at the wing-plate junction, for the wing operating at 4 o incidence. The asymmetric nose-fairing designed for this wing-plate junction is found effective in eliminating the horseshoe vortex and improving the flow quality at the junction. Despite the fairing geometry resulting an increase of wing surface area (by 9.5%) in the wing root region, there was nearly a 6% reduction in the momentum deficit in the plate boundary-layer at x/c of 1.04. However, with the elimination of the horseshoe vortex, the low momentum fluid penetrated into the junction region. This lead to an increase in the boundary-layer shape factor H in the near junction region as compared to the unfaired configuration.

13:00 to 14:00 Break Lunch and Posters
14:00 to 14:15 Prateek Sharma (IISc., India) Multi phase gas around galaxies

Observations show that galaxies like our Milky Way are surrounded by a multiphase (spanning a range of temperatures) “circumglactic medium” (CGM) with the hot (~10​6K) phase confining the cooler (~104K) phase having a much smaller volume filling factor. The structure of this multi phase medium is governed by the inter play of shear driven mixing and cooling which tends to produce distinct temperature phases. Cooling occurs due to atomic transitions in different ion species whose abundance is governed by ionization and recombination processes. Cooling and ionization length / time scales are much smaller compared to the natural scales of this system, making it extremely difficult to simulate the multiphase CGM. There are clear analogies with turbulent combustion and cloud physics which I wish to highlight. This is one of the most exciting recent developments in galaxy formation which deeply involves the physics of turbulent multiphase flows.

14:15 to 14:30 Srinivas Kirthy (IISc., India) Wave packet dispersion and convective/absolutenature of instability in inflectional and non-inflectional flows

In this work, we study spatio-temporal stability characteristics of inflectional and non-inflectional parallel flows and towards this consider the separated boundary layer flow (SBLF) and plane Couette-Poiseuille flow (CPF) as prototypical flowsof each type. It is well-known that the onset of absolute instability in SBLF requires peak reverse flow in the range of 15-20% (Alam and Sandham 2000 J. Fluid Mech.410, 1-28). On the other hand, absolute instability for non-inflectional flows has not been reported to the best of our knowledge. Here we show, for the first time, that CPF becomes absolutely unstable when one of the plates is movedopposite to the bulk motion. Evenarbitrarily small reverse flow is found sufficient to cause absolute instability, provided Reynolds number(Re)is large enough. Further increase in the (negative) plate speed for CPF triggersa transition from absolute instability back to convective instability, however with the wave-packet travelling upstream, i.e., opposite to the bulk fluid motion. This unique transition is not seen inSBLF where absolute instability persists with further increase in reverse flow.Tobetterunderstand this contrasting behaviour, we determine the leadingand trailing edgevelocities of the unstable wave packet following the procedure of Deissler(1987 Phy.Fluids,30,2303-2305).ForCPFwith zero plate speed, we findthat both theedge velocitiesare positive, implying convective instability,whichdecrease andasymptotically approach zero as Re tends to infinity.As a result, asmall negative plate speed enables the trailing-edge velocity to become negative, making the flow absolutely unstable. As the plate speed increases further, the leading-edgevelocityalsobecomes negative, makingthe flow convectively unstable in the upstream direction. For all plate speeds, the difference intheedge velocities, which is a measure of the wave-packetdispersion,initially increases with Re, reaches a maximum and gradually decreases thereafter.However, in the inflectional SBLF, the edge velocitiesand wave-packet dispersionshowaweak dependence on Re, and increasingthereverse flow primarily influencesthe trailing-edge velocity to cross the zero line, whereas the leading-edge velocity remains positive even for large reverse flow. This explains the absence ofthe “reverse” transition from absoluteto convectiveinstabilityinSBLF,which is present inCPF.These results highlight a subtle role of viscosityin triggering and sustaining absolute instability in CPF, but having a weak influence on the dispersion characteristics for SBLF which is dominated by inviscid dynamics.

14:30 to 14:45 Naveen Balakrishna (IISc., India) On late stage mechanisms of transition in round jets
14:45 to 15:00 Anindya Datta (IISc., India) The explicit filtering method for large eddy simulationsof a turbulent premixed flame

Explicit Filtering LES (EFLES) method is used to perform large eddy simula-tions of a turbulent premixed methane-air round jet flame in the TRZ regimeat an equivalence ratio of 0.8 and unburnt gas temperature of 800 K. The nom-inal flow Reynolds number isRe= 1500. EFLES is formally derived from theapproximate deconvolution method for LES computations and has been suc-cessfully applied to non-reacting flow computations in the past. LES resultsare compared with companion DNS results both of which are obtained using areduced 13-step chemical mechanism with species transport to solve for speciesmass fraction fields. LES predicts a slightly shorter mean flame height andthicker flame brush when compared to the DNS due to the lack of fine scaleturbulence and thicker reaction zones in the former. Even so, results for timeaveraged statistics of velocity, temperature and major species mass fractionsfrom LES show remarkably good qualitative and quantitative agreement withDNS results. The use of quasi-steady state assumptions for some intermedi-ate species in the reduced mechanism leads to some quantitative inaccuracy intransported intermediate species fields. Nevertheless, the results obtained areencouraging and show that EFLES is a viable method for reacting flow LEScomputations as well.Keywords:Turbulent premixed flames, Large Eddy Simulation, Explicitfiltering

15:00 to 15:15 Priyanka Maity (ICTS-TIFR, Bengaluru, India) Lagrangian Statistics in rotating turbulent flows

The study of particles suspended in turbulent flows is a topic of active research due to its wide applicabilityin industrial flows, geophysical flows, ans atmospheric flows. Particles suspended in turbulent flows are knowto preferentially sample the flow with spending most of the time in straining regions than in the vorticalregions. Another striking feature of tracers particles suspended in incompressible turbulent flows are known togain energy slowly and loose energy abruptly. This phenomenon, termed as “flight-crash" events [1], canprovide a quantitative measurement of the irreversibility in turbulent flows. The irreversibility was shown toincrease with Taylor-Reynolds number Reλ [1] and Stokes number St [2]. We studied the effect of a constantbackground rotation on these “flight-crash" effects in three-dimensional turbulent flows using direct numericalsimulations (DNS). We observed that rotation generates weak but spatially extended vortical structures in theflow. As a result, in presence of rotation, the particles spends almost equivalent time in straining and vorticalregions. We showed that rotation suppresses the “flight-crash” events in turbulent flows. We quantify the time-irreversibility (Ir = <p3>/ε3) in the flow by computing the third moment of power (p) normalized by the meanenergy dissipation rate (ε). We showed that the irreversibility (Ir) decreases with decreasing Rossby numberRo (increasing rotation rate). The scenario is, however, non-monotonic for heavy particles (St ≠ 0). In case ofheavy particles, the irreversibility, depends on both St and Ro.

15:15 to 16:15 Posters Posters
16:15 to 16:30 Sayantan Majumdar (IISc., India) Jamming under steady-shear in dense particulate suspensions

Under an increasing applied shear stress, viscosity of many dense particulate suspensions increases drastically beyond a stress onset, a phenomenon known as discontinuous shear-thickening (DST). Recent studies point out that some suspensions can transform into a stress induced solid-like shear jammed (SJ) state at high particle volume fractions. SJ state develops a finite yield stress and hence is distinct from a shear-thickened state. Here, we study the steady state shear-thickening behaviour of dense suspensions formed by dispersingcolloidal polystyrene particles in polyethylene glycol. We find that for small values the viscosity of the suspensions as a function of can be well described by Krieger-Dougherty (KD) relation. However, for higher values of applied stress (much larger than the stress onset for shear thickening), KD relation systematically overestimates the measured viscosity,particularly for higher volume fractions.This systematic deviation can be rationalized by the weakening of the sample due to flow induced fractures and failures when the sample is in a solid like SJ state. Using Wyart-Cates model, we propose a method to predict the SJ onset from the steady state rheology measurements. Our results are further supported by in-situ optical imaging of the sample boundary under shear

16:30 to 16:45 Prakriti Pal Choudhury (IISc., India) Circumgalacitc medium-a laboratory for astrophysical fluid dynamics

Multi-wavelength observations detect gas within dark matter halos in various thermodynamic and dynamical states.Broadly, the gas is in three primary dynamical states: firstly, in-falling into the potential well of the halo; secondly,remains within the halo in (quasi)hydrostatic equilibrium until it radiatively cools; thirdly, gets ejected from the halo bystrong outflows (feedback) driven by a central black hole or stars. The gas within the halos and outside the central galaxy(hosting the black hole and stars) is called the circumgalactic medium (CGM), and specifically, intracluster medium(ICM) in the most massive halos (&1014M). Observations indicate that the ICM does not cool as efficiently aspredicted in the theory including pure radiative cooling. Radio, X-ray and Hαobservations now suggest that the centralblack hole is injecting a copious amount of energy in clusters to maintain a rough global thermal balance. Using resultsfrom theory and idealized numerical simulations, I will discuss how locally thermally unstable ICM becomes multiphase(temperature ranges: 104−107K). I will also discuss that the entropy profile of the hot gas gets reset by cooling/feedbackand may modify the nature of multiphase condensation. Lastly, I will mention the ongoing works on the CGM using anew numerical set-up

16:45 to 17:00 Rahul Chajwa (ICTS-TIFR, Bengaluru, India) Waves and instabilities in one dimensional lattice of settling disks

We study experimentally the stokesian sedimentation (Re∼10−4) of a one dimensional latticeof discs in a quasi-two-dimensional geometry with the trajectory of the centres of the disks lying ina plane. We induce initial positional perturbations over a configuration in which the disks are uni-formly spaced with their separation vectors and normals aligned, and perpendicular to gravity. Forvarious perturbation wavenumbers and interparticle separations, we find two classes of behaviour:(i)a transient wave of orientations coupled with number-density fluctuations and (ii) a clumping in-stability resembling that of spheres [J.M. Crowley, J.Fluid Mech. 45, 151 (1971)], decorated withorientations. We construct the equations of motion for displacements and orientations using pair-wise addition of forces and torques [R. Chajwa et al. PRL 122, 224501 (2019)]. Linear stabilityanalysis demarcates a phase boundary between neutrally stable and unstable regimes in the planeof wavenumber and lattice spacing, consistent with our experiments. We predict non-modal growthin this plane, with a critical density of the lattice below which all wavenumbers are asymptoticallystable, showing that orientable particles need not be subject to the inevitable clumping instabilityof spheres.

17:00 to 17:15 Prateek Anand (IISc., India) The dynamics of sedimenting anisotropic particles inturbulence

We examine the dynamics of sub-Kolmogorov anisotropic particles sedimenting throughturbulence. Such systems are ubiquitous in nature, for example, the ice-crystals in Cirrusand mixed-phase clouds. The orientation dynamics of these ice crystals, modelled as prolateand oblate spheroids, plays a crucial role in the planetary greenhouse effect. We perform di-rect numerical simulations(DNS) of such spheroids settling through a homogeneous isotropicturbulent flow field, including the effects of gravity on both the particle translational and ro-tational degrees of freedom. Particle orientation distributions and average settling velocitiesare obtained over a wide range of spheroid aspect ratios, Stokes and Froude numbers. For allcases examined, distributions peak at the broadside-on (to gravity) orientation, and departsignificantly from Gaussianity. These DNS results have been compared against theoreticalpredictions in the inertialess rapid-settling limit, when a particle sediments through a Kol-mogorov eddy much faster than the latter decorrelates. The DNS results deviate from theoryfor Stokes numbers of order unity due to a spatially inhomogeneous particle concentration fieldresulting from a preferential sweeping effect. As a consequence, the time-averaged particle set-tling velocities exceed the orientationally averaged estimates. The spatial inhomogeneity ofthe particle distribution is characterized via a pair-distribution function, allowing an estimateof clustering effects down to the Kolmogorov scale

17:15 to 17:30 Sabarish V N (JNCASR, India) A viscous deformable drop in a linear flow

We study the dynamics of a drop immersed in an ambient fluid subject to a generallinear flow. The main motivation is to study the transport of heat or mass in such dispersemultiphase systems. Such studies have important practical applications ranging from flowswith suspended drops undergoing chemical reactions, vaporisation of atomised fuel jets in aturbulent ambient, which are industrially relevant, to natural phenomena like cloud formation,uptake of nutrients by living cells. We perform Boundary Element simulations (BEM) of asingle drop in a simple shear flow, and compare it to existing theoretical predictions, with theemphasis being on the alteration of the streamline topology around the drop. The simulationsare done over a wide range of viscosity ratios (λ) and the Capillary number (Ca), in theabsence of inertia (Re= 0 ); the Capillary numberCais a dimensionless measure of therelative magnitudes of viscous and interfacial tension forces. For all cases examined, we finda non-zero flux in the flow-gradient plane, and an associated outflux along the vorticity axis.The strength of this flux scales with the Capillary number, and implies a qualitative alterationof the streamline topology due to drop deformation, hitherto unobserved in experiments andunnoticed in earlier computational studies. Our findings have profound implications for theconvective transport from a drop; up until now, the deformation of drop has always beenthought to have only an infinitesimal effect on the transport. An analogous phenomenon ofinertia qualitatively altering the streamline topology, and thereby, enhancing transport rateshad been previously established. The flux associated with the streamlines corresponding tonon-zeroCais found to be opposite to that of the aforementioned inertial case, and the drift issignificantly faster, suggesting deformation-induced convective transport to be the dominantmechanism for large drops in shearing flows.

17:30 to 17:45 Deepak Madival (JNCASR, India) Orientation dynamics of a neutrally-buoyant spheroid of arbitrary aspect-ratio in simple shear flow of weakly viscoelastic fluid