Indian Institute of Science Bangalore
etd@IISc Electronic Theses and Dissertations at Indian Institute of ScienceNot a member yet
6204 research outputs found
Sort by
Electron Microscopy Investigations on Solution Grown Stannous Oxide Nanosheets
No full textIn this thesis, motivated by the possibility of studying rich scientific phenomena using morphology controlled single crystals, experimental observation of new features in the growth and structure of layered oxide materials is explored using stannous oxide (SnO) as a model material. SnO single crystals are grown using a solution method and multiple features of the structure, growth, stability and phase transformation behaviour have been studied experimentally. Growing crystals using solution methods is shown to have specific advantages over other methods and observation of some features are shown to be possible only using multiple electron microscopy techniques. This thesis provides an experimental procedure to study the crystal growth, phase transformation and growth twinning phenomena in a layered oxide material with a well-known sample history and minimal sample preparation for electron microscopy
Design, Fabrication and Characterization of ZnO based Thin Film Schottky Diodes and Transistors
No full textThe thesis focuses on the development of thin film Schottky diodes and thin film transistors (TFTs) based on ZnO. ZnO has been recognized as a promising candidates for the next generation of transparent and flexible electronics for displays. Some of the interesting properties of ZnO include the variation from insulating to semiconducting nature by change of stoichiometry, the relative low toxicity enabling its use in edible materials, the presence of a reasonably high electron mobility and its high transmission to visible light. All of these properties have increased interest for the development of ZnO-TFTs and diodes.
This work focuses on process development of thin film Schottky diodes( Al-ZnO-Ag) and transistors(Al-ZnO-ZrO2). The Schottky diodes were developed with thermally evaporated Aluminium ohmic contact and silver Schottky contact. The fabricated diodes had cut-in voltage between 1-2 V with mean reverse saturation current of 1.0 x 10^-7 A and an excellent rectification ratio of 10^6. Thin film transistors were developed with thermally evaporated Aluminium contacts for Gate, Source and Drain. Zinc oxide was used as semiconductor channel material. For process development of thin film transistors, Zinc oxide was used as semiconductor and a transparent thin film with transmittance of 83.45 % at 450 nm was deposited using DC Reactive sputtering of zinc in oxygen ambient of 1 x 10^-3 mbar. The optical bandgap was found to be around 3.15 eV. ZrO2 was selected as Gate dielectric because of its high dielectric constant, wide band gap and excellent chemical and thermal stability. The ZrO2 thin film was deposited by DC reactive sputtering in an oxygen ambient of 1.5 x 10^-3 mbar. The maximum drain to source current was found to be 25.45 mA and maximum leakage gate current was found to be 0.22 mA
An Evaluation of Basic Protection Mechanisms in Financial Apps on Mobile Devices
No full textThis thesis concerns the robustness of security checks in financial mobile applications (or simply
financial apps). The best practices recommended by OWASP for developing such apps demand
that developers include several checks in these apps, such as detection of running on a rooted
device, certificate checks, and so on. Ideally, these checks must be introduced in a sophisticated
way, and must not be locatable through trivial static analysis, so that attackers cannot bypass
them trivially. In this thesis, we conduct a large-scale study focused on financial apps on the
Android platform and determine the robustness of these checks. Our study shows that among
the apps with at least one security check, > 50% of such apps at least one check can be trivially
bypassed. Some of such financial apps we considered have installation counts exceeding 100
million from Google Play. We believe that the results of our study can guide developers of these
apps in inserting security checks in a more robust fashion.Department of Science and Technology, Govt. of India
Development of Novel Deep Learning Methods for Fast-MRI: Anatomical Image Reconstruction to Quantitative Imaging
No full textIn medical imaging, the task of estimating interpretable anatomical images from raw scanner data - based on underlying physical principles - is known as an "inverse problem". The solution to such inverse problems can be as simple as inverse Fourier transform for Magnetic Resonance Imaging (MRI). However, MRI is inherently slow due to the requirement of filling in the "k-space data". One of the popular ways of reducing the scan time is to use highly under-sampled data (collecting only a few samples of k-space data). Fast-MRI methods have found greater utility in dynamic imaging (3D+time), like Dynamic Contrast-Enhanced (DCE) MRI for cancer diagnosis and MR angiography. The acceleration in data acquisition time can be achieved using mathematical algorithms that incorporate these techniques' physical principles. This makes the inverse problem more challenging. Data driven methods based on deep learning (DL) have been able to provide promising results with few questions to be addressed like data dependency, lack of interpretability and lack of uncertainty quantification. This thesis work proposes physics-based DL algorithms that work with less training data and are more interpretable with the utilization of a physics-based forward model. The developed networks are also robust to data perturbations. Specifically, this thesis work addressed two problems related to Fast-MRI, with the first one concerning anatomical image reconstruction and the other focusing on quantitative imaging.
Anatomical image reconstruction: In this part, a generic deep learning-based MR image reconstruction model (named SpiNet) was proposed that can enforce any Schatten p-norm regularization with 0 < p <= 2, where the p can be learnt (or fixed) based on the problem at hand. Model-based deep learning architecture for solving inverse problems consists of two parts, a deep learning based denoiser and an iterative data consistency solver. The former has either L2 norm or L1 norm enforced on it, which are convex and can be easily minimized. This thesis work proposes a method to enforce any p norm on the noise prior where 0 < p ≤ 2. This is achieved by using the Majorization–Minimization algorithm, which upper bounds the cost function with a convex function, and thus can be easily minimized. The proposed SpiNet has the capability to work for a fixed p or it can learn p based on the data. The network was tested for solving the inverse problem of reconstructing magnetic resonance (MR) images from undersampled k space data and the results were compared with state-of-the-art model-based deep learning architecture (MoDL) which enforces L2 norm along with other compressive sensing-based algorithms. This comparison between the current state of the art methods and proposed SpiNet was performed for undersampling rates (R) of 2×, 4×, 6×, 8×, 12×, 16×, and 20×. Multiple figures of merit such as PSNR, SSIM, and NRMSE were utilized in this comparison. A two-tailed t-test was performed for all undersampling rates and for all metrices for proving the superior performance of the proposed SpiNet. The results indicate that for all undersampling rates, the proposed SpiNet shows higher PSNR and SSIM and lower NRMSE than other state-of-the-art methods. However, for low undersampling rates of 2× and 4×, there is no significant difference in performance of proposed SpiNet and other state-of-the-art methods in terms of PSNR and NRMSE. This can be expected as the learnt p value is close to 2 (norm enforced by other methods). For higher undersampling rates grater than 6×, SpiNet significantly outperforms current state-of-the-art method in all metrices with improvement as high as 4 dB in PSNR and 0.05 points in SSIM.
Quantitative imaging: In this second part, this work focussed on estimating the permeability parameters from highly undersampled Dynamic Contrast-Enhanced (DCE) MR images and consists of two investigations. In the first investigation, the performance of iterative direct and indirect parametric reconstruction methods with indirect deep learning-based reconstruction methods in estimating tracer-kinetic parameters from highly undersampled DCE-MR Imaging breast data and provide a systematic comparison of the same. The investigations concluded that deep learning-based indirect techniques perform at par with direct estimation techniques for lower undersampling rates in the breast DCE-MR imaging. At higher undersampling rates, they are not able to provide much needed generalization. Direct estimation techniques are able to provide more accurate results than both deep learning- and parametric-based indirect methods in these high undersampling scenarios. The second investigation is of development novel physics-based DL scheme for permeability parameter estimation called Greybox - an amalgamation of DL (black box) and iterative techniques (white box). This algorithm is invariant to the undersampling rate and tested this algorithm for brain, breast and prostate cancer patients. Additionally, this thesis work also proposed a pure DL based architecture for direct estimation of permeability parameters. Unlike existing architectures, this network has been invariant to the spatial and temporal size of input data.
Deep learning based methods have shown promise for solving inverse problems associated with Fast-MRI. This thesis work shown that deep learning methods can also provide much needed quantitative accuracy, other than the obvious added advantage of being computationally efficient, for making MRI the most preferred imaging modality for quantitative imaging with applications in oncology and neuroimaging. The methods proposed here are able to generalise across anatomical structures and data sets, showing the versatility and making them appealing even in clinical settings.Prime Minister's Research Fellowshi
Effect of Length Scale on High Temperature Mechanical Behavior of Sn-Cu Joints: A Mechanics and Material Science Based Treatment
No full textWith the ongoing miniaturization of microelectronic devices, the size of compliant solder in microscale solder joints has significantly reduced, proportion of brittle phases has increased, and the microscale joints have become highly constrained to deform because of their geometry and stiffness mismatch with substrates. Due to the varied nature of microstructure, a mere understanding of the mechanical behaviour of bulk alloys cannot be directly extended to judiciously predict the same in miniature joints. Accordingly, the aim of this work is to develop a comprehensive understanding of the effect of microstructure and size of Sn based joints on the mechanical behavior and microstructure from bulk specimens to microscale joints. This was first addressed by investigating the joint size dependence of the tensile properties of Sn-Cu joints. Maximum strength increased as the joint size reduced, and the mode of failure changed from necking in thick joints to constrained necking with cavitation in microscale Sn-Cu joints and solder-IMC failure in miniature Sn-Ag-Cu/Cu joints. The cause of this tensile strengthening was captured by crystal plasticity (CP) modeling along with existing analytical models to capture the size dependence of the tensile strength. Subsequently the effect of length scale on the creep properties of the Sn-Cu joints was investigated. This was addressed by first evaluating the creep behaviour of bulk Sn and Sn-Ag-Cu solder alloys over a range of temperature to compute the activation energy, QC, and stress exponent, n. The creep rate decreased with increasing Ag content. The creep mechanism was dislocation climb controlled by core diffusion at T150 oC with QC and n changing from 55 kJ/mol and 7 to 100 kJ/mol and 5. Subsequently the size of deformable Sn and solder was reduced by constraining metal layers of different size (from 1.4 mm to ~170 µm) and aspect ratio between Cu substrates as joints. The secondary creep rate decreased by three orders of magnitude with an order of magnitude decrease in joint size at same stress. However, no change in creep mechanism was evident in the joints. Using Finite Element Analysis this creep strengthening can be partly attributed to geometric constraints imposed by Cu which reduces the effective stress and increases the triaxiality in joints. While FE results were in close agreement with experiment in thick joints and bulk specimen, the former overpredicted the creep rate of small joints due to difference in microstructure between bulk Sn which has multiple grains and miniature joints having 2-3 grains, as confirmed by electron backscattered diffraction. This microstructural effect was captured by dislocation based crystal plasticity modeling from which it was evident that orientation anisotropy of Sn and constraints imposed by substrates on dislocation motion can lead to additional strengthening in small joints and reduce the prefactor B in Norton power law. Subsequently, a unified model was developed to quantify both these effects and predict the creep rate of arbitrary joints. In the tertiary creep stage bulk specimens exhibited strain localization by necking in pure Sn along with cavitation in precipitate containing Sn-Ag-Cu alloys. Moreover, the extent of necking and cavitation in Sn-Cu joints was sensitive to the joint size. A combined necking and cavitation based creep failure model was developed and the effect of initial geometry of instability, creep stress exponent and cavity fraction on strain localization was analyzed. The model showed that the strain at the onset of complete localization in neck will be independent of stress in pure Sn whereas it decreases with stress in Sn-Ag-Cu alloy and the predictions were found to qualitatively agree with experiment. The model also was adapted to capture the joint size dependence of tertiary creep due to necking and cavitation in the Sn-Cu and SAC-Cu joints. The model predictions and experiments showed a decrease in strain accumulated in tertiary stage with decrease in joint size, as the strain due to necking reduced and the strain due to cavitation increased from bulk Sn and solders to the very thin joints
Microscale mechanical behaviour of ceramic matrix composites considering processing e ects
No full textThe current work explores the phenomenon of microscale damage mechanism driven quasi-
ductile behaviour of C/BN/SiC ceramic matrix composite (CMC) at the microscale focusing
on process-microstructure-property correlations. C/BN/SiC minicomposites have been fabri-
cated by chemical vapour infiltration (CVI) with varying interphase thicknesses and constituent
volume fractions by varying the interphase (BN) and matrix (SiC) in filtration durations. The
effect of processing durations on the resulting microstructure, tensile response and damage
mechanisms up to and during ultimate failure of CMC minicomposites have been obtained ex-
perimentally that highlight the significant infuence of processing duration on the tensile and
failure behaviour of CMC minicomposites.
Processing induced micro-scale matrix porosity in the fabricated minicomposites has been
characterized by X-ray micro-computed tomography. Effective elastic properties in the presence
of matrix micro-pores have been obtained by a two-step numerical homogenization approach
that includes the statistical distributions of pore parameters obtained from experimental char-
acterization. A variation of the approach has been utilized to investigate the severity of pores
with respect to their location and orientation relative to the fiber reinforcement.
A probabilistic progressive damage modeling approach has been proposed to predict the
tensile response of CMC minicomposites considering the microstructural information from fab-
ricated minicomposites. The highlight of the proposed numerical approach is the development
of a 3 phase shear lag model to better approximate matrix crack driven stress transfer in the
presence of an interphase between the ber and the matrix. The in
uence of volume frac-
tions, constituent properties and interfacial properties on the mechanical behavior of CMC
minicomposites have been presented.
The presented approaches and results provide an insight into the processing-microstructure-
tensile response relationship and the e ect of processes induced defects on the tensile response
in CMCs. Additionally, modeling approaches have been proposed for predicting the tensile
response of CMCs at the microscale considering processing induced defects
Droplet Isothermal Amplification For Nucleic Acid Quantification
No full textNucleic acid quantification (NAQ) is extensively employed for gene expression analysis, monitoring viral loads, detecting rare or dysfunctional cells, and assessing treatment regimes. The gold standard, quantitative polymerase chain reaction (qPCR), and the recent alternative, droplet digital PCR (ddPCR), provide accurate quantification of nucleic acids (NA). Albeit the requirement of thermal cycling and separate platforms for droplet generation, NA amplification, and signal detection, in the case of ddPCR increases the assay complexity and time, limiting its broad applicability.
In this work, we have developed an integrated droplet isothermal amplification-based NAQ (idNAQ) platform that enables facile and fast NAQ with a large dynamic range. First, we adapted the isothermal amplification method, Recombinase Polymerase Amplification (RPA), for NAQ. We demonstrate a fast (• 40 minutes) semi-quantitative RPA (qRPA) assay with the endpoint intensity ratio (EIR) for DNA quantification with a 6-log order range. Since the EIR model estimates the amplicon levels at the end of the reaction, real-time monitoring of the amplification reaction (unlike in the case of qPCR) is no longer required. With qRPA, we demonstrate viral load detection from the serum of dengue-infected patients with comparable performance to qPCR. The later section discusses the translation of the qRPA NAQ to a microfluidic droplet format. Droplet RPA (dRPA) displays similar kinetics to the bulk reaction suggesting successful optimization of droplet conditions for RPA. dRPA in the low concentration regime follows Poisson distribution that enables digital quantification as in the case of ddPCR. On the other hand, at a higher starting concentration of DNA (non-digital regime, >10 DNA per droplet), the RPA amplification in droplets exhibits heterogeneous intensity puncta due to rapid amplification and incomplete mixing leading to the formation of DNA ‘amplification globules’. We use a supervised machine, learning-based regression model with these intensity features as inputs to accurately predict the target concentration of up to 10^5 molecules per droplet. Combining these two modalities of dRPA yields a dynamic range of >7 log orders of concentration that are comparable to qPCR. Finally, we demonstrate the successful integration of all unit operations onto a single microfluidic device for droplet RPA and quantification. Different microfluidic designs were optimized for monodisperse droplet generation and image acquisition from a large incubation area that allowed the successful implementation of quantitative dRPA in a single device
On the development of sensible heat storage for concentrated solar power applications: Thermo-fluid management and materials
No full textSensible heat storages have extensive use in thermal energy deployment, including concentrated
solar power (CSP) applications. Usually, CSP pants demand various techno-economic features
in sensible heat storage, such as low-cost, high-capacity, efficiency, and ease of operation. These
requirements demand investigations to assess and develop novel strategies to improve the efficacy
of sensible heat thermal energy storage (TES) technology. Accordingly, the present study
focuses on thermo-fluid management and material characterization for stratified TES.
Computational fluid dynamics simulations were employed to analyze near-inlet thermal blending
of hot and cold heat transfer fluid (HTF), molten salt, for a single-tank sensible heat TES
system. Accordingly, a hemispherical diffuser is developed. In addition, a mathematical index
is proposed to quantify the degree of thermal stratification. Further, experiments were conducted
for thermosyphon charging of single-tank stratified storage including both continuous
and pulsatile charging at low (150 °C) and high (250 and 300 °C) temperatures. Dowtherm-A
oil was used as the HTF, and the thermal expansion of HTF was accommodated in an expansion
tank via two different designs (top and bottom connections from the storage tank to the
expansion tank).
From a materials viewpoint, high specific heat capacity (CP ) is essential to improve the energy
density of the storage; which can be improved by adding nanoparticles to molten salt. However,
the literatures show both increment and decrement in CP . Since difficulties are associated with
identifying explicit relations between molten salts and nanoparticles due to complex molecular
interactions, we inquired whether there are common patterns/clusters in the nanofluid samples
reported in earlier studies by employing unsupervised machine learning methods: Hierarchical
cluster analysis (HCA) and Principal component analysis (PCA). Finally, a comparative analysis
is presented to capture the measurement variability in nanofluid samples under random
sampling. In this analysis, the DSC test is employed on small-sized batches (< 10 mg) and the
T-history method on large-sized batches (∼ 20g), and the CP values of both tests are compared
using a nonparametric statistical test, Mann-Whitney U Test
Discrete particulate description of elastic structures undergoing geometrically nonlinear deformation and dynamic particle interaction
No full textThe mechanical behaviour of deformable bodies in a particulate environment has been an area of increasing interest across a wide spectrum of systems and scale. A composite ensemble of deformable structures and discrete particles involves coupling of component responses, large displacements of structures, and multiple dynamic interactions that lead to inherent contact nonlinearity. To describe these structures and their interactions with particles, we apply a particle simulation approach based on the discrete element method (DEM). There exist alternative frameworks too such as continuum modelling with techniques like finite element method (FEM), or a combination of continuum and discrete modelling, or lumped modelling with mass-spring systems. Owing to the convenience and robustness provided by a single approach, this thesis aims to develop a single framework with the discrete modelling approach.
The mechanical behaviour of particulate models of slender elastic continua is first validated with their analytical or FEM counterparts, and then particle-structure interactions are considered. We develop elastically deformable particulate models of straight beams and shallow arches. We evaluate the geometrically nonlinear response of particulate beams under a variety of static, dynamic, and impact loading scenarios. We also model particulate arches and assess their ability to exhibit two force-free equilibrium states, namely bistability. To illustrate the utility of these particulate representations, we first consider a case study of an undulating beam in a particle medium. The dynamic beam-particle interactions propel the beam within the medium, resembling the self-propulsion of reptiles in granular environments. In another case study, we take up a relatively sparse environment of mobile particles and oscillating cantilever beams. The interplay between particles and beams is shown to drive particles for capture. We also demonstrate particle-arch interactions in bistable mechanisms that result in particle gripping and trapping. We draw insights from factors that regulate the governing dynamics of such coupled phenomena.
Next, we model particulate thin films that undergo deflections in linear and geometrically nonlinear regimes and describe both plate-like and membrane-like behaviours. A notable instance of this particulate perspective of thin films occurs in the context of microscopic biological material such as cells and their organelle. We present in this context a discrete particulate description for the nucleus of a biological cell. A three-dimensional model that incorporates the nuclear envelope and chromatin-containing nucleoplasm is developed and subjected to micropipette aspiration. Our work on particulate systems is implemented within Altair EDEM, a commercial DEM software. While most available DEM packages, including EDEM, provide a ready-to-use interface for the modelling and analysis of granular and bulk materials, they lack similar modules for particulate structures. The preprocessing stage thus involves substantive customization to build algorithms for particle generation, contact physics, external couplings, and parameter definition appropriate to our studies. By customizing this process, we utilize the graphical interface capabilities of EDEM to simulate, readjust, and visualize the analysis of structures under applied forces and particle interaction.
Taken together, the studies in this thesis facilitate a comprehensive investigation of the particulate approach’s efficacy to model a variety of deformable structures, capture geometric nonlinearity in their response, and simulate the interaction dynamics of coupled particle-structure systems
Automated methods of natural resource mapping with Remote Sensing Big data in Hadoop MapReduce framework
No full textFor several decades, remote sensing (RS) tools have provided platforms for the large-scale exploration of natural resources across the planetary bodies of our solar system. In the context of Indian remote sensing, mineral resources are being explored, and mangrove resources are being monitored towards a sustainable socio-economic structure and coastal eco-system, respectively, by utilising several remote analytical techniques. However, RS technologies and the corresponding data analytics have made a vast paradigm shift, which eventually has produced “RS Big data” in our scientific world of large-scale remote analysis. Consequently, the current practices in remote sensing need a systematic improvisation of data analytics to provide a real-time, accurate and feasible remote exploration of the RS Big data. Towards this, the improvement of corresponding scientific analysis has opened up new opportunities and research perspectives for both academia and industry in remote sensing. In this favour, different automated methods are proposed in the Hadoop MapReduce framework as a part of this thesis aiming to develop both decisive and time-efficient remote analysis under the RS Big data environment. This thesis studies the remote exploration of various surface types covering the mineralogy and mangrove regions, respectively, as two significant applications in natural resource mapping. Before starting, the reliability and outreach of RS Big data analysis in the Hadoop MapReduce framework are also assessed in the laboratory environment. In this thesis, each proposed automated method is validated first in the single node analysis as a standalone process for individual RS applications. Then the corresponding MapReduce designs of the proposed methods make them scaled to conduct the distributed analysis in a pseudo-distributed Hadoop architecture for a prototype RS Big data environment in this thesis.
In particular, a “per-pixel” mapping of the mineralised belt is conducted with a proposition of Extreme Learning Machine (ELM)-based scaled-ML algorithm in the Hadoop MapReduce framework by addressing the primary challenge because of impurity in the representative spectra of an observed pixel. To an extent, the same mineralogical province is explored with a proposition of a fraction cover mapping model in the Hadoop MapReduce framework by addressing the primary challenge due to the spectral variation of pure mineral spectra within an observed pixel. These mineralogical explorations on Earth utilise airborne-based hyperspectral imagery, whereas mineralogical explorations on Moon utilise spaceborne-based hyperspectral lunar imagery in this thesis. An automated mineralogical anomaly detection method identifies the prominent lunar mineral occurrences by addressing the consequences of space weathering on lunar exposures. On the other side, the spaceborne-based active remote sensing of polarimetric Earth imagery is utilised for land cover classification over the mangrove region in the Hadoop MapReduce framework. The land features of fully polarimetric (FP) and compact polarimetric (CP) observations are explored with a proposition of Active learning Multi-Layered Perceptron (AMLP) by addressing the primary challenge due to the uncertainties in class labelling. The robustness, stability, and generalisation of all proposed shallow neural networks of single hidden layered ML models are analysed for varietal informative data classification. In fact, the advancements in methodology and architecture support each other in attaining a better remote analysis with less computational automated methods in this thesis.
Some of the crucial findings of this thesis are as follows: For a reliable and generalised mineral mapping, the perturbed/mixed spectra of hydrothermal minerals are required to be mapped along with the pure spectra of hydrothermal minerals. Further, the fraction cover mapping of hydrothermal minerals should address the spectral variation of pure spectra and the underlying physics of spectral mixing to get a reliable and accurate fractional contribution of minerals. In contrast to Earth mineralogy, the automated lunar mineral exploration needs to identify the potential mineralogical map of the lunar surface because of the space weathering effect. On the other hand, the underlying physics behind the polarimetric synthetic aperture radar (SAR) remote sensing plays a vital role in better discrimination of land features within the mangrove regions. The inherent data parallelism technique of the Hadoop architecture simply makes the analytical algorithm scaled and time-efficient, which can be extended for real-time Big data environments even with other MapReduce frameworks. In conclusion, even shallow learning of an automated method can provide an efficient real-time analysis of the RS Big data prototype if the physical constraints or prior physics-based insights of remote observations are undertaken. It is evident in this thesis that such consideration makes the prototype RS Big data analysis more reliable, accurate, scalable, automated and widely acceptable under varietal remote sensing environments. In summary, this thesis builds a bridge between academia and industry to provide new directional research on RS Big data analysis in making a better real-time futuristic plan for the natural resource management of any country like India