Reshma Pagoti

Topic - Deep Unrolled Networks for Signal and Image Denoising via Sparse Signal Recovery

Presentation Date: 3rd January 2025
Time:
Location: Seminar Hall, New Building, Indian Institute of Petroleum and Energy

Summary of the presentation – Recently, Deep Neural Networks (DNNs) have been developed as a popular approach for addressing the Compressed Sensing problem. Typically, in DNN, a neural network is trained on a designated dataset using a specified loss function and then evaluated on previously unseen test data. Despite their effectiveness, DNNs often lack interpretability and generalizability, frequently relying on trial and error. This reliance results in complex architectures with high training costs and substantial data requirements. To address this, a technique known as algorithm unrolling has been developed. Motivated by this, in this presentation, we discuss the unrolling of the Generalized Orthogonal Matching Pursuit algorithm for signal and image denoising tasks.

Geetha Singh

Topic - Development of Deep Learning Approach to Simulate Fracture Propagation in Rock

Presentation Date: 3rd January 2025
Time:
Location: Seminar Hall, New Building, Indian Institute of Petroleum and Energy

Summary of the presentation – The prediction of deformation mechanisms in geological materials is critical for excavation activities, including tunnels, slopes, oil and gas extraction, and mining. Extensive research and development have focused on developing robust computational methods, which traditionally involve numerical approximations of the governing equations to determine displacement or velocity fields under prescribed boundary conditions but face challenges with increasing problem complexity, particularly for non-linear material behavior or data integration. Recently, Physics-Informed Neural Networks (PINNs) have emerged as a promising alternative, embedding governing laws within neural network architectures to solve PDEs and address forward and inverse problems. This work focuses on implementing the PINN model to solve various problems in geomechanics.

Diwakar Patel

Topic - Assessment of Sustainable Materials for Wastewater Remediation

Presentation Date: 27th Dec. 2024
Time:
Location: Seminar Hall, New Building, Indian Institute of Petroleum and Energy

Summary of the presentation – The need for clean water is rising due to population growth and intensive industrialization. Industries like textiles, pulp and paper, leather, and tanneries release dye-containing effluent into water sources and are responsible for around 20% of total global wastewater. This research examines the elimination of dyes from industrial dye effluent by adsorption methods utilizing carbon nano onions (CNOs) produced from waste frying oil.

L. Karunakar Reddy

Topic - Numerical Evaluation of PCM Based Energy Storage System

Presentation Date: 13th Dec. 2024

Summary of the presentation:
Latent heat thermal energy storage (LHTES) units utilize phase change materials (PCMs) to effectively store thermal energy and address the challenges of solar energy intermittency and instability. Heat transfer characteristics largely influence the thermo-fluidic efficiency of the LHTES unit during phase transitions. To improve the typically low heat transfer capabilities of PCMs, high-conductivity metal foams are introduced to enhance bulk heat conduction. This study focuses on optimizing the parameters of composite PCM (CPCM) embedded in the LHTES unit, specifically examining metal foam height (H), angle (θ), and porosity (ε), which primarily affect the performance of the CPCM. A series of Taguchi-based orthogonal experiments were conducted to analyze the effects of these variables. The results indicate that metal foam height has a notable influence on the CPCM performance, significantly reducing melt fraction variations caused by changes in cross-sectional geometry.

Rajeswar Das

Topic - Effects of Thermal Treatment on Physical Properties and Deformational Behaviour of Rocks

Presentation Date: 13th Dec. 2024

Summary of the presentation:
The influence of temperature on the physical-mechanical properties of rock has drawn considerable attention in recent years, especially in various geotechnical projects such as geothermal energy extraction, underground coal gasification, shale gas production, nuclear waste disposal, enhanced oil recovery, drilling and blasting, and tunnel fires. Basaltic terrain, which is found both in oceanic and continental crust, holds potential for geothermal energy production and underground nuclear waste storage due to its high thermal conductivity, specific heat capacity, hardness, and strength. This study aims to evaluate the effects of heating and cooling on the physical properties and deformational behavior of basaltic rocks collected from the Rajahmundry Traps, Andhra Pradesh, India. The findings of this study contribute to understanding thermal damage evolution in basalt at elevated temperatures and support the evaluation of safety and stability in deep rock masses for various underground geoengineering applications confined to basaltic terrain.

Pradyumna Pradhan

Topic - The Role of Uni-modular Tight Frames in Image Compression and Reconstruction

Presentation Date: 6 th Dec. 2024
Time:
Location: Seminar Hall, New Building, Indian Institute of Petroleum and Energy

Summary of the presentation – A highly redundant incoherent unimodular tight frame (UMTF) is highly desirable in various fields, including quantum communications and signal processing. The incorporation of a larger number of orthogonal blocks into such a UMTF provides significant advantages, particularly in applications like image compression and wireless communications. While UMTFs with orthogonal blocks are valued for their low computational cost, existing approaches for their construction are limited. Current methods often suffer from constraints such as a small number of orthogonal blocks and high computational complexity. This work addresses these limitations by developing a highly redundant UMTF with a large number of orthogonal blocks and reduced computational complexity compared to existing methods.

Ankit Rawal

Topic - Antifouling membrane for industrial application

Presentation Date: 6 th Dec. 2024
Time:
Location: Seminar Hall, New Building, Indian Institute of Petroleum and Energy

Summary of the presentation – Low-temperature plasma surface modification using a PECVD reactor is employed to enhance the surface properties of polysulfone (PSf) membranes, focusing on improving their antifouling performance for industrial wastewater treatment. This technique enables precise control over surface characteristics, such as hydrophilicity and roughness, to mitigate fouling and improve operational longevity. Untreated and plasma-treated membranes are compared systematically using both batch cells and crossflow filtration systems to assess their performance under varying operational conditions. Key parameters such as separation efficiency, throughput, and rejection rates are analyzed to evaluate the impact of plasma treatment. The findings aim to optimize membrane functionality, offering sustainable and efficient solutions for practical applications in wastewater management..

Sahanawaz Parvez (H&S)

Topic- Inhibitors development against human 5-LOX

Presentation date -25.10.2024
Time: 3:00 PM – 4:00 PM
Location: Seminar Hall, New Building, Indian Institute of Petroleum and Energy

Summary of the presentation – Human Lipoxygenases (LOX) are enzymes which catalyze the oxygenation of polyenoic fatty acids. 5-LOX is a LOX isoform which converts arachidonic acids (AA) into leukotrienes (LTA 4 ). Leukotrienes have important roles in inflammation progression in asthma, cardiovascular diseases, allergic rhinitis, atopic dermatitis, inflammatory bowel disease, rheumatoid arthritis, psoriasis and many cancers and its inhibition has great therapeutic value. Thus, in order to suppress leukotrienes formation for the management of such diseases, the intrinsic details of the structure of 5-LOX are crucial for the development of 5- LOX inhibitors. In our study, we have deciphered the role of important amino acids in the allosteric binding site by using molecular dynamics simulation which can be applied in novel potent drug discovery against 5-LOX.