Resource
2019 EN
Jayanta Bera · Satyajit Sahu
Two-dimensional transition metal dichalcogenides show great potential aspromising thermoelectric materials due to their lower dimensionality, theunique density of states and quantum confinement of carriers. The effect ofmechanical strain on the thermoelectric performances of monolayer WS 2 has beeninvestigated using density functional theory associated with semiclassicalBoltzmann transport theory. The variation of Seebeck coefficient and band gapwith applied strain has followed the same type of trend. For n-type materialthe relaxation time scaled power factor(S 2 {\sigma}/{\tau}) increases by theapplication of compressive strain whereas for p- type material it increaseswith the application of tensile strain. A 77% increase in the power factor hasbeen observed for the n-type material by the application of uniaxialcompressive strain. A decrease in lattice thermal conductivity with theincrease in temperature causes an almost 40% increase in ZT product underapplied uniaxial compressive strain. From the study, it is observed thatuniaxial compressive strain is more effective among all types of strain toenhance the thermoelectric performance of monolayer WS 2 . Such strain inducedenhancement of thermoelectric properties in monolayer WS 2 could open a newwindow for the fabrication of high-quality thermoelectric devices.
Resource
2019 EN
Preeti Sahu · Janice Kang · Gonca Erdemci-Tandogan
+1 more
The rheology of biological tissues is important for their function, and wewould like to better understand how single cells control global tissueproperties such as tissue fluidity. A confluent tissue can fluidize when cellsdiffuse by executing a series of cell rearrangements, or T1 transitions. In adisordered 2D vertex model, the tissue fluidizes when the T1 energy barriersdisappear as the target shape index approaches a critical value ($s^*_{0} \sim3.81$), and the shear modulus describing the linear response also vanishes atthis same critical point. However, the ordered ground states of 2D vertexmodels become linearly unstable at a lower value of the target shape index(3.72) [1,2]. We investigate whether the ground states of the 2D vertex modelare fluid-like or solid-like between 3.72 and 3.81 $-$ does the "equation ofstate" for these systems have two branches, like glassy particulate matter, oronly one? Using four-cell and many-cell numerical simulations, we demonstratethat for a hexagonal ground state, T1 energy barriers only vanish at $\sim3.81$, indicating that ordered systems have the same critical point asdisordered systems. We also develop a simple geometric argument that correctlypredicts how non-linear stabilization disappears at $s^*_{0}$ in orderedsystems.
Resource
2019 EN
Sunil Kumar Sahu · Fenia Christopoulou · Makoto Miwa
+1 more
Inter-sentence relation extraction deals with a number of complex semanticrelationships in documents, which require local, non-local, syntactic andsemantic dependencies. Existing methods do not fully exploit such dependencies.We present a novel inter-sentence relation extraction model that builds alabelled edge graph convolutional neural network model on a document-levelgraph. The graph is constructed using various inter- and intra-sentencedependencies to capture local and non-local dependency information. In order topredict the relation of an entity pair, we utilise multi-instance learning withbi-affine pairwise scoring. Experimental results show that our model achievescomparable performance to the state-of-the-art neural models on twobiochemistry datasets. Our analysis shows that all the types in the graph areeffective for inter-sentence relation extraction.
Resource
2019 EN
Sarira Sahu
Blazars are a subclass of AGN and flaring in multi-TeV gamma-ray seems to bethe major activity in high energy blazars a subgrup of blazars. Flaring is alsounpredictable and switches between quiescent and active states involvingdifferent time scales and fluxes. While in some high energy blazars a strongtemporal correlation between X-ray and multi-TeV gamma-ray has been observed,outbursts in some other have no low energy counterparts and explanation of suchextreme activity needs to be addressed through different mechanisms as it isnot understood well. The extragalactic background light (EBL) plays animportant role in the observation of these high energy gamma-rays as itattenuates through pair production of electron-positron and also changes thespectral shape of the high energy photons. In the context of the photohadronicmodel and taking EBL correction into account, flaring can be explained verywell. In a series of papers we have developed this model to explain multi-TeVflaring events form many blazars. Here in this review, the photohadronic modelis discussed and applied to explain the multi-TeV flaring from nearby highenergy blazars: Markarian 421, Markarian 501 and 1ES1959+650.
Resource
2019 EN
Surendra Kumar Soni · Pavan Kumar Kirar · Pankaj Kolhe
+1 more
We experimentally investigate the deformation and breakup of dropletsinteracting with an oblique continuous air stream. A high-speed imaging systemis employed to record the trajectories and topological changes of the dropletsof different liquids. The droplet size, the orientation of the air nozzle tothe horizontal and fluid properties (surface tension and viscosity) are variedto study different breakup modes. We found that droplet possessing initialmomentum prior to entering the continuous air stream exhibits a variation inthe required Weber number for the vibrational to the bag breakup transitionwith a change in the angle of the air stream. The critical Weber numbers$(We_{cr})$ for the bag-type breakup are obtained as a function of theE\"{o}tv\"{o}s number $(Eo)$, angle of inclination of the air stream $(\alpha)$and the Ohnesorge number $(Oh)$. It is found that although the droplet followsa rectilinear motion initially that transforms to a curvilinear motion at latertimes when the droplet undergoes topological changes. The apparent accelerationof the droplet and its size influence the critical Weber number for the bagbreakup mode. The departure from the cross-flow arrangement shows a sharpdecrease in the critical Weber number for the bag breakup which asymptoticallyreaches to a value associated with the in-line (opposed) flow configuration forthe droplet breakup.
Resource
2019 EN
Pradeep Gurrala · Pallavi Katre · Saravanan Balusamy
+2 more
We experimentally investigate the evaporation dynamics of sessile droplets ofa fixed volume consisting of different compositions of ethanol-water binarymixture at different substrate temperatures (T_s). At T_s=25oC, we observepinned-stage linear evaporation for pure droplets, but a binary dropletundergoes two distinct evaporation stages: an early pinned stage and a laterreceding stage. In the binary droplet, the more volatile ethanol, evaporatesfaster leading to a nonlinear trend in the evaporation process at the earlystage. The phenomenon observed in the present study at T_s=25oC is similar tothat presented by previous researchers at room temperature. More interestingdynamics is observed in the evaporation process of a binary droplet at anelevated substrate temperature. We found that the lifetime of the dropletexhibits a non-monotonic trend with the increase in ethanol concentration inthe binary mixture, which {can be attributed to} the non-ideal behaviour ofwater-ethanol binary mixtures. Increasing T_s decreases the lifetime of the(50\% ethanol + 50 \% water) binary droplet in a logarithmic scale. For thiscomposition, at T_s=60oC, we observed an early spreading stage, an intermediatepinned stage and a late receding stage of evaporation. Unlike T_s=25oC, at theearly times of the evaporation process, the contact angle of the droplet ofpure water at T_s=60oC is greater than 90. Late stage interfacial instabilityand even droplet break-up are observed for some (though not all) binary mixturecompositions. The evaporation dynamics for different compositions at T_s=60oCexhibit a self-similar trend. Finally, the evaporation rates of pure and binarydroplets at different substrate temperatures are compared against a theoreticalmodel developed for pure and binary mixture droplets.
Resource
2019 EN
Siddhant Sahu · Manoj Kumar Lenka · Pankaj Kumar Sa
We inspect all the deep learning based solutions and provide holisticunderstanding of various architectures that have evolved over the past fewyears to solve blind deblurring. The introductory work used deep learning toestimate some features of the blur kernel and then moved onto predicting theblur kernel entirely, which converts the problem into non-blind deblurring. Therecent state of the art techniques are end to end, i.e., they don't estimatethe blur kernel rather try to estimate the latent sharp image directly from theblurred image. The benchmarking PSNR and SSIM values on standard datasets ofGOPRO and Kohler using various architectures are also provided.
Resource
2019 EN
Vishesh Vatsal · C. Barath · J. Yogeshwaran
+7 more
TeamIndus' lunar logistics vision includes multiple lunar missions to meetrequirements of science, commercial and efforts towards global exploration. Thefirst mission is slated for launch in 2020. The prime objective is todemonstrate autonomous precision lunar landing, and Surface Exploration Roverto collect data on the vicinity of the landing site. TeamIndus has developedvarious technologies towards lowering the access barrier to the lunar surface.This paper shall provide an overview of design of lander GNC system. The designof the GNC system has been described after concluding studies on sensor andactuator configurations. Frugal design approach is followed in the selection ofGNC hardware. The paper describes the constraints for the orbital maneuvers andthe lunar descent strategy. Various aspects of the GNC design of autonomouslunar descent maneuver: timeline of events, guidance, inertial and opticalterrain-relative navigation schemes are described. The GNC software descriptionfocuses on system architecture, modes of operation, and core elements of theGNC software. The GNC algorithms have been tested using Monte-Carlo simulationsand Processor-in-Loop runs. The paper concludes with a summary of keyrisk-mitigation studies for soft landing.
Resource
2019 EN
David P. Bennett · Rachel Akeson · Thomas Barclay
+22 more
WFIRST is NASA's first flagship mission with pre-defined core scienceprograms to study dark energy and perform a statistical census of wide orbitexoplanets with a gravitational microlensing survey. Together, these programsare expected to use more than half of the prime mission observing time.Previously, only smaller, PI-led missions have had core programs that used sucha large fraction of the observing time, and in many cases, the data from thesePI-led missions was reserved for the PI's science team for a proprietary periodthat allowed the PI's team to make most of the major discoveries from the data.Such a procedure is not appropriate for a flagship mission, which shouldprovide science opportunities to the entire astronomy community. For thisreason, there will be no proprietary period for WFIRST data, but we argue thata larger effort to make WFIRST science accessible to the astronomy community isneeded. We propose a plan to enhance community involvement in the WFIRSTexoplanet microlensing survey in two different ways. First, we propose a set ofhigh level data products that will enable astronomers without detailedmicrolensing expertise access to the statistical implications of the WFIRSTexoplanet microlensing survey data. And second, we propose the formation of aWFIRST Exoplanet Microlensing Community Science Team that will open upparticipation in the development of the WFIRST exoplanet microlensing survey tothe general astronomy community in collaboration for the NASA selected scienceteam, which will have the responsibility to provide most of the high level dataproducts. This community science team will be open to volunteers, but membersshould also have the opportunity to apply for funding.
Resource
2019 EN
Minyoung Kim · Yuting Wang · Pritish Sahu
+1 more
We propose a family of novel hierarchical Bayesian deep auto-encoder modelscapable of identifying disentangled factors of variability in data. While manyrecent attempts at factor disentanglement have focused on sophisticatedlearning objectives within the VAE framework, their choice of a standard normalas the latent factor prior is both suboptimal and detrimental to performance.Our key observation is that the disentangled latent variables responsible formajor sources of variability, the relevant factors, can be more appropriatelymodeled using long-tail distributions. The typical Gaussian priors are, on theother hand, better suited for modeling of nuisance factors. Motivated by this,we extend the VAE to a hierarchical Bayesian model by introducing hyper-priorson the variances of Gaussian latent priors, mimicking an infinite mixture,while maintaining tractable learning and inference of the traditional VAEs.This analysis signifies the importance of partitioning and treating in adifferent manner the latent dimensions corresponding to relevant factors andnuisances. Our proposed models, dubbed Bayes-Factor-VAEs, are shown tooutperform existing methods both quantitatively and qualitatively in terms oflatent disentanglement across several challenging benchmark tasks.