arXiv subjects

Arjun Sharma

Publications and source records attributed to Arjun Sharma.

At least 19 recordsLinked to original sources

arXiv2026

An inertial slender-body theory

We present a fully inertial slender-body theory (SBT) that incorporates the effect of fluid inertia on the scale of the length (the "outer" region) as well as the characteristic diameter (the "inner" region) of a steadily translating slender particle. This is achieved by matching the solution of the quasi-two-dimensional full Navier-Stokes equations in the inner region to an outer solution that consists of a superposition of a solution of the linearized Navier-Stokes equations driven by a line of forces and a potential flow solution driven by a line distribution of sources and source dipoles. The drag and lift forces result from the distribution of Oseen force singularities. These Oseenlets also predominantly govern the torque at small Reynolds numbers and large aspect ratios. However, the potential flow singularities play a crucial role in yielding a torque that grows with increasing Reynolds number at large Reynolds numbers and finite aspect ratios. By comparing the forces and torque on the steadily translating particle with those obtained from a finite difference Navier-Stokes solution, we demonstrate the accuracy of the resulting inertial SBT for $\mathrm{Re}_D$ up to 10, where $\mathrm{Re}_D$ is the Reynolds number based on the smallest dimension, i.e., the characteristic cross-sectional diameter of the slender particle.

physics.flu-dyn↗

arXiv2026

Data-informed lifting line theory

We present a data-driven framework that extends the predictive capability of classical lifting-line theory (LLT) to a wider aerodynamic regime by incorporating higher-fidelity aerodynamic data from panel method simulations. A neural network architecture with a convolutional layer followed by fully connected layers is developed, comprising two parallel subnetworks to separately process spanwise collocation points and global geometric/aerodynamic inputs such as angle of attack, chord, twist, airfoil distribution, and sweep. Among several configurations tested, this architecture is most effective in learning corrections to LLT outputs. The trained model captures higher-order three-dimensional effects in spanwise lift and drag distributions in regimes where LLT is inaccurate, such as low aspect ratios and high sweep, and generalizes well to wing configurations outside both the LLT regime and the training data range. The method retains LLT's computational efficiency, enabling integration into aerodynamic optimization loops and early-stage aircraft design studies. This approach offers a practical path for embedding high-fidelity corrections into low-order methods and may be extended to other aerodynamic prediction tasks, such as propeller performance.

physics.flu-dyn↗

arXiv2026

Simulating the swimming motion of a flagellated bacterium in a microstructured bio-fluid

We develop a numerical framework to simulate the locomotion of a flagellated bacterium with a spheroidal head (such as Escherichia coli) in biological fluids like mucus, which are entangled polymer solutions exhibiting elasto-viscoplastic (EVP) rheology and porous microstructure. To account for the scale disparity between the large bacterial head and the slender flagellar bundle, whose thickness is comparable to the pore size, we employ a two-fluid model in which the bundle directly drives the solvent and exchanges momentum with the polymer phase via drag proportional to their relative velocity. The numerical implementation combines a finite-difference discretization of the two-fluid equations with a slender-body theory (SBT) to model flagellar forcing. A key observation is that the coupled mass and momentum equations for these inertialess flows, together with SBT, are linear in the pressure and velocity fields and in the force distribution along the flagellar bundle. By treating the polymer stress as a body force, we decompose the flow field and hydrodynamic moments into three additive contributions: kinematic (motion), flagellar forcing, and polymer stress. This decomposition allows several components of the flow to be precomputed and enables the determination of swimming velocity via a resistivity formulation driven by polymer-induced forces, which greatly improves computational efficiency during transient calculations of the polymer stress and the resulting flow. We validate the method and use it to analyze how polymer microstructure and its interactions with the bacterial head and tail affect motility in complex biofluids.

physics.flu-dyn↗

arXiv2025

Stress analysis of dilute particle suspensions in non-Newtonian fluids with efficient evaluation in the weakly non-Newtonian limit

We present a semi-analytical framework to compute the suspension stress in dilute particle-laden non-Newtonian fluids, separating Newtonian and non-Newtonian contributions. The ensemble-averaged stress includes both the particle-induced non-Newtonian stress (PINNS) and an interaction stresslet arising from surface tractions due to the non-Newtonian stress and its induced Newtonian flow. Using a generalized reciprocal theorem, we express this interaction stresslet entirely in terms of the non-Newtonian stress, for a general constitutive model. For weakly non-Newtonian fluids, a regular perturbation expansion combined with the method of characteristics yields all leading-order stress contributions from the Newtonian velocity field alone, avoiding the need to solve coupled partial differential equations. This generalizes the method of Koch et al. (Phys. Rev. Fluids 1, 013301 (2016)) beyond polymeric fluids to any weakly non-Newtonian medium driven by velocity and its gradients. We apply the method to two systems: (i) spheres suspended in a fluid of smaller spheroids, where the interaction stress becomes negative for sufficiently anisotropic shapes due to orientation misalignment of the spheroids; and (ii) suspensions in weakly anisotropic nematic liquid crystals. In the latter, assuming a uniform director field fixed by an external field, PINNS vanishes while interaction stresslets remain, either opposing or enhancing background anisotropic stress. These results demonstrate the utility of our framework in capturing first-order particle-microstructure interactions across a broad class of non-Newtonian fluids.

cond-mat.soft↗

arXiv2025

Viscosity variation in fluid flows across scale

A wide range of natural and engineered fluid flows exhibit spatial or temporal viscosity variations, spanning scales from microbial locomotion to planetary mantle convection. These variations introduce qualitatively new physical mechanisms absent in constant-viscosity flows. This review surveys such phenomena across scales. In low Reynolds number (Stokes) flows, viscosity gradients couple translation and rotation, enabling novel particle responses to uniform forcing-- mechanisms that microorganisms may exploit. In shear flows, viscosity variation alters base flow profiles and breaks symmetries, modifying stability and transition dynamics. At high Reynolds numbers, stratification fundamentally changes the singular perturbation structure governing energy production, enhancing or suppressing canonical instabilities and introducing new ones. Viscosity variation also affects nonnormal growth and nonlinear interactions that drive transition to turbulence. While laminar and fully developed turbulence have been extensively studied, transitional processes remain poorly understood in variable-viscosity flows. In turbulent regimes, viscosity variation impacts jets, wall-bounded flows, and mixing layers. At geophysical scales, incorporating eddy viscosity stratification in climate models may improve predictions, while in Earth's mantle, viscosity contrasts drive large-scale convection and geological evolution. Particle-laden flows, common across contexts, can generate effective viscosity stratification through inhomogeneous loading. Throughout, we highlight cases where viscosity variation alters flow behavior qualitatively, and point to open questions. This review aims to guide graduate students and researchers toward tractable, cross-disciplinary problems.

physics.flu-dyn↗

arXiv2025

Extensional rheology of dilute suspensions of spheres in polymeric liquids

The extensional rheology of dilute suspensions of spheres in viscoelastic or polymeric liquids is studied computationally. At low polymer concentration (c) and Deborah number (De), a wake of highly stretched polymers forms downstream of the particles due to larger local velocity gradients than the imposed flow, indicated by a positive deviation in local De. This increases the suspension's extensional viscosity with time and De for De less than 0.5. When De exceeds 0.5 (the coil-stretch transition), the fully stretched polymers from the far field collapse in regions with lower local velocity gradients around the particle's stagnation points, reducing suspension viscosity relative to the polymer-only liquid. The interaction between local flow and polymers intensifies with increasing c. Highly stretched polymers impede local flow, reducing local De, while it increases in regions with collapsed polymers. Initially, increasing c aligns local De and polymer stretch with far-field values, diminishing particle-polymer interaction effects. However, beyond a certain c, a new mechanism emerges. At low c, fluid three particle radii upstream exhibits increased local De, stretching polymers beyond their undisturbed state. As c increases, this deviation becomes negative, collapsing polymers and resulting in increasingly negative stress from particle-polymer interactions at large De and time. At high c, this negative interaction stress scales as c squared, surpassing the linear increase in polymer stress, making dilute sphere suspensions more effective at reducing the viscosity of viscoelastic liquids at larger De and c.

physics.flu-dyn↗

arXiv2025

Sedimentation and shear-induced dynamics of spheroids in fluids with spatial viscosity variations

A generalized reciprocal theorem is used to relate the force and torque induced on a particle in an inertia-less fluid with small variation in viscosity to integrals involving Stokes flow fields and the spatial dependence of viscosity. These resistivity expressions are analytically evaluated using spheroidal harmonics and then used to obtain the mobility of the spheroid during sedimentation, and in linear flows, of a fluid with linear viscosity stratification. The coupling between the rotational and translational motion induced by stratification rotates the spheroid's centerline, creating a variety of rotational and translational dynamics dependent upon the particle's aspect ratio, $\kappa$, and the component of the stratification unit vector in the gravity direction, $d_g$. Spheroids with $0.55\lessapprox\kappa\lessapprox2.0$ exhibit the largest variety of settling behaviors. Interestingly, this range covers most microplastics and typical microorganisms. One of the modes include a stable orientation dependent only on $\kappa$ and $d_g$, but independent of initial orientation, thus allowing for the potential control of settling angles and sedimentation rates. In a simple shear flow, cross-streamline migration occurs due to the stratification-induced force generated on the particle. Similarly, a particle no longer stays at the stagnation point of a uniaxial extensional flow. While fully analytical results are obtained for spheroids, numerical simulations provide a source of validation. These simulations also provide additional insights into the stratification-induced force- and torque-producing mechanisms through the stratification-induced stress, which is not accessed in the reciprocal theorem-based analytical calculations.

physics.flu-dyn↗

arXiv2024

Wing Optimisation for a tractor propeller driven Micro Aerial Vehicle

This paper describes an investigation of the possible benefits from wing optimisation in improving the performance of Micro Air Vehicles (MAVs). As an example we study the Avion (3.64 kg mass, 1.60 m span), being designed at the CSIR National Aerospace Laboratories (NAL), Bengaluru. The optimisation is first carried out using the methodology described by Rakshith \emph{et al.} (using an in\textendash house software PROWING), developed for large transport aircraft, with certain modifications to adapt the code to the special features of the MAV. The chief among such features is the use of low Reynolds number aerofoils with significantly different aerodynamic characteristics on a small MAV. These characteristics are taken from test data when available, and/or estimated by the XFOIL code of Drela. A total of 8 optimisation cases are studied for the purpose, leading to 6 different options for new wing planforms (and associated twist distributions along the wing span) with an improved performance. It is found that the improvements in drag coefficient using the PROWING code are about 5%. However, by allowing the operating lift coefficient $C_L$ to float within a specified range, drag bucket characteristics of the Eppler E423 aerofoil used on Avion can be exploited to improve the endurance, which is a major performance parameter for Avion. Thus, compared to the control wing $W_0$ (with operating point at $C_L$ =0.7) used in the preliminary design, permitting a variation of $C_L$ over a range of $\pm$ 10% is shown to enhance the endurance of wing $W_4$ by 18.6%, and of wing $W_{6}$ with a permitted $C_L$ range of $\pm$ 50% by 39.2%. Apart from the philosophy of seeking optimal operating conditions for a given configuration, the advantages of optimising design parameters such as washout of a simple wing proposed in the preliminary design stage, is also demonstrated.

eess.SY↗

arXiv2024

Positive and unlabelled machine learning reveals new fast radio burst repeater candidates

Fast radio bursts (FRBs) are astronomical radio transients of unknown origin. A minority of FRBs have been observed to originate from repeating sources, and it is unknown which apparent one-off bursts are hidden repeaters. Recent studies increasingly suggest that there are intrinsic physical differences between repeating and non-repeating FRBs. Previous research has used machine learning classification techniques to identify apparent non-repeaters with repeater characteristics, whose sky positions would be ideal targets for future observation campaigns. However, these methods have not sufficiently accounted for the positive and unlabelled (PU) nature of the data, wherein true labels are only available for repeaters. Modified techniques that do not inadvertently learn properties of hidden repeaters as characteristic of non-repeaters are likely to identify additional repeater candidates with greater accuracy. We present in this paper the first known attempt at applying PU-specific machine learning techniques to study FRBs. We train an ensemble of five PU-specific classifiers on the available data and use them to identify 66 repeater candidates in burst data from the CHIME/FRB collaboration, 18 of which were not identified with the use of machine learning classifiers in past research. Our results additionally support repeaters and non-repeaters having intrinsically different physical properties, particularly spectral index, frequency width, and burst width. This work additionally opens new possibilities to study repeating and non-repeating FRBs using the framework of PU learning.

astro-ph.HE↗

arXiv2023

Finite difference method in prolate spheroidal coordinates for freely suspended spheroidal particles in linear flows of viscous and viscoelastic fluids

A finite difference scheme is used to develop a numerical method to solve the flow of an unbounded viscoelastic fluid with zero to moderate inertia around a prolate spheroidal particle. The equations are written in prolate spheroidal coordinates, and the shape of the particle is exactly resolved as one of the coordinate surfaces representing the inner boundary of the computational domain. As the prolate spheroidal grid is naturally clustered near the particle surface, good resolution is obtained in the regions where the gradients of relevant flow variables are most significant. This coordinate system also allows large domain sizes with a reasonable number of mesh points to simulate unbounded fluid around a particle. Changing the aspect ratio of the inner computational boundary enables simulations of different particle shapes ranging from a sphere to a slender fiber. Numerical studies of the latter particle shape allow testing of slender body theories. The mass and momentum equations are solved with a Schur complement approach allowing us to solve the zero inertia case necessary to isolate the viscoelastic effects. The singularities associated with the coordinate system are overcome using L'Hopital's rule. A straightforward imposition of conditions representing a time-varying combination of linear flows on the outer boundary allows us to study various flows with the same computational domain geometry. {For the special but important case of zero fluid and particle inertia we obtain a novel formulation that satisfies the force- and torque-free constraint in an iteration-free manner.} The numerical method is demonstrated for various flows of Newtonian and viscoelastic fluids around spheres and spheroids (including those with large aspect ratio). Good agreement is demonstrated with existing theoretical and numerical results.

physics.flu-dyn↗

arXiv2023

Rotation of a fiber in simple shear flow of a dilute polymer solution

The motion of a freely rotating prolate spheroid in a simple shear flow of a dilute polymeric solution is examined in the limit of large particle aspect ratio, $\kappa$. A regular perturbation expansion in the polymer concentration, $c$, a generalized reciprocal theorem, and slender body theory to represent the velocity field of a Newtonian fluid around the spheroid are used to obtain the $\mathcal{O}(c)$ correction to the particle's orientational dynamics. The resulting dynamical system predicts a range of orientational behaviors qualitatively dependent upon $c\cdot De$ ($De$ is the imposed shear rate times the polymer relaxation time) and $\kappa$ and quantitatively on $c$. At a small but finite $c\cdot De$, the particle spirals towards a limit cycle near the vorticity axis for all initial conditions. Upon increasing $\kappa$, the limit cycle becomes smaller. Thus, ultimately the particle undergoes a periodic motion around and at a small angle from the vorticity axis. At moderate $c\cdot De$, a particle starting near the flow-gradient plane departs it monotonically instead of spirally, as this plane (a limit cycle at smaller $c\cdot De$) obtains a saddle and an unstable node. The former is close to the flow direction. Upon further increasing $c\cdot De$, the saddle-node changes to a stable node. Therefore, depending upon the initial condition, a particle may either approach a periodic orbit near the vorticity axis or obtain a stable orientation near the flow direction. Upon further increasing $c\cdot De$, the limit cycle near the vorticity axis vanishes, and the particle aligns with the flow direction for all starting orientations.

physics.flu-dyn↗

arXiv2023

Towards Autonomous and Safe Last-mile Deliveries with AI-augmented Self-driving Delivery Robots

In addition to its crucial impact on customer satisfaction, last-mile delivery (LMD) is notorious for being the most time-consuming and costly stage of the shipping process. Pressing environmental concerns combined with the recent surge of e-commerce sales have sparked renewed interest in automation and electrification of last-mile logistics. To address the hurdles faced by existing robotic couriers, this paper introduces a customer-centric and safety-conscious LMD system for small urban communities based on AI-assisted autonomous delivery robots. The presented framework enables end-to-end automation and optimization of the logistic process while catering for real-world imposed operational uncertainties, clients' preferred time schedules, and safety of pedestrians. To this end, the integrated optimization component is modeled as a robust variant of the Cumulative Capacitated Vehicle Routing Problem with Time Windows, where routes are constructed under uncertain travel times with an objective to minimize the total latency of deliveries (i.e., the overall waiting time of customers, which can negatively affect their satisfaction). We demonstrate the proposed LMD system's utility through real-world trials in a university campus with a single robotic courier. Implementation aspects as well as the findings and practical insights gained from the deployment are discussed in detail. Lastly, we round up the contributions with numerical simulations to investigate the scalability of the developed mathematical formulation with respect to the number of robotic vehicles and customers.

cs.RO↗

arXiv2022

Steady state extensional rheology of a dilute suspension of spheres in a dilute polymer solution

We investigate the steady-state extensional rheology of a dilute suspension of spherical particles in a dilute polymer solution. For a particle-free polymeric fluid, in addition to the solvent viscosity, the extensional viscosity due to the polymers, $\mu^\text{poly}$, contributes to the total non-dimensionalized extensional viscosity $1+\mu^\text{poly}$. When a small volume fraction, $\phi$, of spheres is added to a polymeric fluid, the stress is altered by the Einstein viscosity of 2.5$\phi$ and two additional stress contributions: the interaction stresslet and the particle-induced polymer stress (PIPS). The net interaction stress is positive at lower Deborah numbers (product of extension rate and polymer relaxation time), $De\lesssim0.5$, and negative at large $De$. Relative to undisturbed flow, the presence of spheres in uniaxial extensional flow creates larger and smaller local stretching regions. Below the coil-stretch transition ($De<0.5$), the polymers far from the particles are in a coiled state, and a wake of stretched polymers forms downstream of the particle as they are stretched by the large stretching regions around the particle. This leads to a positive interaction stresslet (surface) and the PIPS (stretched wake). Beyond the coil-stretch transition, polymers far from the particle are highly stretched, but they collapse closer to the coiled state as they arrive at the low-stretching regions near the particle surface. Therefore, a negative PIPS results from the regions of collapsed polymers. When $De\gtrsim0.6$, the changes in extensional viscosity from the interaction stresslet and the PIPS are $\phi\mu^\text{poly}$ and approximately -1.85$\phi\mu^\text{poly}$, respectively. At large $De$, the polymer extensional viscosity, $\mu^\text{poly}$, is large. Therefore, adding particles reduces the extensional viscosity of the suspension ($(2.5-0.85\mu^\text{poly})\phi<0$).

cond-mat.soft↗

arXiv2021

Chest ImaGenome Dataset for Clinical Reasoning

Despite the progress in automatic detection of radiologic findings from chest X-ray (CXR) images in recent years, a quantitative evaluation of the explainability of these models is hampered by the lack of locally labeled datasets for different findings. With the exception of a few expert-labeled small-scale datasets for specific findings, such as pneumonia and pneumothorax, most of the CXR deep learning models to date are trained on global "weak" labels extracted from text reports, or trained via a joint image and unstructured text learning strategy. Inspired by the Visual Genome effort in the computer vision community, we constructed the first Chest ImaGenome dataset with a scene graph data structure to describe $242,072$ images. Local annotations are automatically produced using a joint rule-based natural language processing (NLP) and atlas-based bounding box detection pipeline. Through a radiologist constructed CXR ontology, the annotations for each CXR are connected as an anatomy-centered scene graph, useful for image-level reasoning and multimodal fusion applications. Overall, we provide: i) $1,256$ combinations of relation annotations between $29$ CXR anatomical locations (objects with bounding box coordinates) and their attributes, structured as a scene graph per image, ii) over $670,000$ localized comparison relations (for improved, worsened, or no change) between the anatomical locations across sequential exams, as well as ii) a manually annotated gold standard scene graph dataset from $500$ unique patients.

cs.CV↗

arXiv2021

AnaXNet: Anatomy Aware Multi-label Finding Classification in Chest X-ray

Radiologists usually observe anatomical regions of chest X-ray images as well as the overall image before making a decision. However, most existing deep learning models only look at the entire X-ray image for classification, failing to utilize important anatomical information. In this paper, we propose a novel multi-label chest X-ray classification model that accurately classifies the image finding and also localizes the findings to their correct anatomical regions. Specifically, our model consists of two modules, the detection module and the anatomical dependency module. The latter utilizes graph convolutional networks, which enable our model to learn not only the label dependency but also the relationship between the anatomical regions in the chest X-ray. We further utilize a method to efficiently create an adjacency matrix for the anatomical regions using the correlation of the label across the different regions. Detailed experiments and analysis of our results show the effectiveness of our method when compared to the current state-of-the-art multi-label chest X-ray image classification methods while also providing accurate location information.

cs.CV↗

arXiv2020

Creation and Validation of a Chest X-Ray Dataset with Eye-tracking and Report Dictation for AI Development

We developed a rich dataset of Chest X-Ray (CXR) images to assist investigators in artificial intelligence. The data were collected using an eye tracking system while a radiologist reviewed and reported on 1,083 CXR images. The dataset contains the following aligned data: CXR image, transcribed radiology report text, radiologist's dictation audio and eye gaze coordinates data. We hope this dataset can contribute to various areas of research particularly towards explainable and multimodal deep learning / machine learning methods. Furthermore, investigators in disease classification and localization, automated radiology report generation, and human-machine interaction can benefit from these data. We report deep learning experiments that utilize the attention maps produced by eye gaze dataset to show the potential utility of this data.

cs.CV↗

arXiv2020

Spatio-temporal relationships between rainfall and convective clouds during Indian Monsoon through a discrete lens

The Indian monsoon, a multi-variable process causing heavy rains during June-September every year, is very heterogeneous in space and time. We study the relationship between rainfall and Outgoing Longwave Radiation (OLR, convective cloud cover) for monsoon between 2004-2010. To identify, classify and visualize spatial patterns of rainfall and OLR we use a discrete and spatio-temporally coherent representation of the data, created using a statistical model based on Markov Random Field. Our approach clusters the days with similar spatial distributions of rainfall and OLR into a small number of spatial patterns. We find that eight daily spatial patterns each in rainfall and OLR, and seven joint patterns of rainfall and OLR, describe over 90\% of all days. Through these patterns, we find that OLR generally has a strong negative correlation with precipitation, but with significant spatial variations. In particular, peninsular India (except west coast) is under significant convective cloud cover over a majority of days but remains rainless. We also find that much of the monsoon rainfall co-occurs with low OLR, but some amount of rainfall in Eastern and North-western India in June occurs on OLR days, presumably from shallow clouds. To study day-to-day variations of both quantities, we identify spatial patterns in the temporal gradients computed from the observations. We find that changes in convective cloud activity across India most commonly occur due to the establishment of a north-south OLR gradient which persists for 1-2 days and shifts the convective cloud cover from light to deep or vice versa. Such changes are also accompanied by changes in the spatial distribution of precipitation. The present work thus provides a highly reduced description of the complex spatial patterns and their day-to-day variations, and could form a useful tool for future simplified descriptions of this process.

physics.ao-ph↗

arXiv2020

Looking in the Right place for Anomalies: Explainable AI through Automatic Location Learning

Deep learning has now become the de facto approach to the recognition of anomalies in medical imaging. Their 'black box' way of classifying medical images into anomaly labels poses problems for their acceptance, particularly with clinicians. Current explainable AI methods offer justifications through visualizations such as heat maps but cannot guarantee that the network is focusing on the relevant image region fully containing the anomaly. In this paper, we develop an approach to explainable AI in which the anomaly is assured to be overlapping the expected location when present. This is made possible by automatically extracting location-specific labels from textual reports and learning the association of expected locations to labels using a hybrid combination of Bi-Directional Long Short-Term Memory Recurrent Neural Networks (Bi-LSTM) and DenseNet-121. Use of this expected location to bias the subsequent attention-guided inference network based on ResNet101 results in the isolation of the anomaly at the expected location when present. The method is evaluated on a large chest X-ray dataset.

cs.CV↗