http://dspace.dtu.ac.in:8080/jspui/handle/123456789/42 2026-10-24T02:59:57Z 2026-10-24T02:59:57Z GUPTA, SHAILESH KUMAR Srivastava, Amit Kumar (SUPERVISOR) http://dspace.dtu.ac.in:8080/jspui/handle/repository/23405 2026-10-16T06:37:01Z 2026-10-01T00:00:00Z

Title: EFFECT OF HETEROGENEITY ON FLOW CHARACTERISTICS IN POROUS MEDIA Authors: GUPTA, SHAILESH KUMAR; Srivastava, Amit Kumar (SUPERVISOR) Abstract: Heterogeneous and stratified porous media exhibit complex flow behaviour because variations in soil composition, pore structure, layer arrangement, and orientation directly influence seepage pathways and hydraulic conductivity. Reliable assessment of these effects is important for geotechnical, groundwater, waste-containment, drainage, and environmental engineering applications. The present study investigates the combined influence of chemical modification, microbiological treatment, and layer inclination on the horizontal hydraulic conductivity of heterogeneous porous media. Laboratory experiments were carried out using a specially designed horizontal permeameter with overall dimensions of 0.8 m × 0.5 m × 0.5 m and a 0.6 m long soil column. Two inorganic sandy-silt soils collected from the Yamuna alluvial region of Delhi were arranged in stratified form to represent heterogeneous porous conditions. Fly ash was selected as the chemical admixture and incorporated into the upper soil layer at 5%, 10%, 15%, 20%, and 25% by dry weight. Bacillus subtilis was used as the microbiological agent at bacterial cell densities of 10⁶, 10⁷, and 10⁸ cells/mL. Experimental testing was performed at layer inclination angles varying from 0° to 10° to quantify the independent and combined effects of chemical content, bacterial concentration, and specimen orientation on horizontal hydraulic conductivity. The experimental results demonstrated that the hydraulic response of the stratified media was strongly affected by all three governing variables. In untreated specimens, horizontal hydraulic conductivity increased with increasing layer inclination, indicating improved continuity of the preferential flow path along the oriented layers. The addition of fly ash consistently lowered hydraulic conductivity, and the degree of reduction increased with increasing fly ash dosage. At 0° inclination, reductions of approximately 10.83%, 27.23%, 40.07%, 56.04%, and 65.94% were obtained for 5%, 10%, 15%, 20%, and 25% fly ash, respectively. This behaviour is associated with pore filling, densification, and restriction of interconnected void spaces. Microbiological treatment also produced a marked decrease in hydraulic conductivity. Higher bacterial cell densities generated greater reductions because bacterial growth, extracellular polymeric substances, and pore occupation progressively restricted the available seepage channels. The combined use of fly ash and Bacillus subtilis produced the strongest hydraulic modification, demonstrating a synergistic effect between physical pore filling and biological clogging. vi The maximum reduction in horizontal hydraulic conductivity was approximately 91.33% at 0° inclination for the combined condition of 25% fly ash and 10⁸ cells/mL bacterial concentration. Although conductivity generally increased with inclination angle, treated specimens remained considerably less permeable than untreated stratified media throughout the tested range. In addition to the experimental investigation, Support Vector Regression (SVR) was used to model horizontal hydraulic conductivity as a function of soil characteristics, fly ash content, bacterial cell density, layer inclination, and number of layers. The model reproduced the nonlinear experimental behaviour with strong predictive performance and low error, demonstrating the usefulness of machine-learning techniques for representing complex heterogeneous flow systems. Overall, the study confirms that the hydraulic behaviour of stratified porous media can be substantially modified through controlled chemical and microbiological interventions. The findings provide a practical basis for reducing seepage and improving hydraulic performance in applications related to groundwater protection, engineered barriers, waste containment, soil improvement, and hydraulic infrastructure.

2026-10-01T00:00:00Z CHOUKSEY, ARTI Dev, Nirendra (SUPERVISOR) http://dspace.dtu.ac.in:8080/jspui/handle/repository/23401 2026-10-16T06:36:24Z 2026-10-01T00:00:00Z

Title: MECHANICAL AND MICROSTRUCTURAL CHARACTERIZATION OF LIME SLUDGE- METAKAOLIN BLENDED CONCRETE COMPOSITE Authors: CHOUKSEY, ARTI; Dev, Nirendra (SUPERVISOR) Abstract: In this study, two by-products, lime sludge (LS) and Metakaolin (MK), were used as partial replacements for cement to enhance the mechanical, durability and microstructural properties of concrete. MK is a reactive aluminosilicate pozzolan, whereas LS is a calcium-rich by-product of the paper industry. The investigation was carried out in two linked phases. In the first phase, M25 grade concrete was prepared with LS and MK at replacement levels of 0–15% each (a 4 × 4 factorial set of 16 mixes), and the effect of these materials was studied in terms of workability, density, compressive strength, split tensile strength, flexural strength, water absorption, acid resistance and chloride ion penetration. Regression analysis was performed to examine the relationships among strength, density and water absorption, and scanning electron microscopy and energy-dispersive spectroscopy were employed to examine the hydration products and the microstructure. The mix with 10% LS and 10% MK (C80M10L10) performed best among all the LS–MK mixes. Compared with the control mix with 0% LS and 0% MK, it attained about 23% and 26% higher compressive strength at 28 and 90 days, 13–15% higher split tensile strength and 13 15% higher flexural strength. Its water absorption was 20% lower (6.4% versus 8.0%), its mass loss after 12 weeks in 5% HCl was 8% against 10% for the control, and its 56-day RCPT charge was about 64% lower (800 C versus 2,200 C). Regression analysis showed positive but moderate correlations between density and the strength properties (R² = 0.24–0.36) and a stronger inverse correlation between water absorption and density (R² = 0.54). The complete 4 × 4 factorial matrix of MK LS blends corresponds to a Taguchi L16 (4²) orthogonal array, with MK and LS as the two control factors at four levels (0, 5, 10 and 15%). ANOVA of the signal-to noise ratios showed that MK was the dominant factor, contributing about 41–89% of the total variation in compressive strength, water absorption, RCPT charge and acid induced mass loss, while LS contributed about 8–31%; because each mix was tested as a single run, the MK × LS interaction could not be separated from the error term. The cost analysis showed that replacing 20% of the cement with 10% MK and 10% LS increased the material cost of the M25 concrete by about 10.2%, mainly because of the higher price of MK, as lime sludge adds very little cost; this marginal increase is justified by the improvements in strength, durability, microstructure and xvi sustainability. Overall, the mix with 10% MK and 10% LS is technically and economically feasible and suitable for durable and eco-friendly concrete applications. Microstructural analysis indicated a denser matrix and a refined interfacial transition zone (ITZ) for C80M10L10, and the EDS-derived Ca/Si ratios were consistent with the formation of additional C–S–H/C–A–S–H gel, although these phases were inferred rather than directly identified. In the second phase, the performance of M40 grade concrete incorporating MK and calcined lime sludge (CLS) at 0–20% each was investigated. Workability decreased with increasing replacement owing to the higher specific surface areas of MK and CLS, whereas the mechanical properties improved at moderate replacement levels. The mix containing 15% MK and 10% CLS (M15CL10) exhibited the highest compressive strength (55.2 MPa at 90 days against 49.0 MPa for the control), the highest split tensile and flexural strengths and the lowest water absorption, indicating a synergistic effect between the pozzolanic activity of MK and the calcium-rich composition of CLS; beyond this replacement level, strength decreased and water absorption increased. Quadratic response surface models fitted to the CLS–MK data confirmed 15% MK and 10% CLS as the optimum (desirability 0.95). A cradle-to-gate life cycle assessment showed that the optimum mixes reduce the global warming potential by about 18% (C80M10L10) and 21% (M15CL10) relative to their respective controls. Therefore, this study establishes LS–MK and CLS–MK blended concretes as technically feasible and more sustainable alternatives to conventional concrete, contributing to waste valorisation and a reduced carbon footprint.

2026-10-01T00:00:00Z KUMAR, ISHAAN SARKAR, RAJU (SUPERVISOR) http://dspace.dtu.ac.in:8080/jspui/handle/repository/23396 2026-10-16T06:35:45Z 2026-03-01T00:00:00Z

Title: LANDSLIDE SUSCEPTIBILITY ASSESMENT IN SIKKIM USING FREQUENCY RATIO AND SHANNON ENTROPY APPROACH Authors: KUMAR, ISHAAN; SARKAR, RAJU (SUPERVISOR) Abstract: Landslide Susceptibility Assessment (LSA) is crucial for understanding and mitigating landslide risks, especially in vulnerable regions like Sikkim, where complex topography, diverse geological formations, and significant precipitation contribute to instability. This study integrates geospatial analysis with statistical modelling techniques to evaluate the influence of 10 landslide conditioning factors (LCFs), including aspect, rainfall, curvature, elevation, hillshade, roughness, slope, distance to stream and topographic wetness index (TWI) and normalized difference vegetation index (NDVI). Using a landslide inventory of 677 recorded events and GIS-derived thematic layers were used to apply two statistical models: Frequency Ratio (FR) and Shannon Entropy (SE). Landslide inventory is created by dividing the total landslide points into training data (80%) and testing data (20%). The vulnerability map is then created using the 2 models and a comparison is done between them. Then, the map created by the above-mentioned 2 models is validated with the help of a testing data set using the AUC (Area under curve) of the ROC (Receiver Operating Characteristics) curve. The resulting susceptibility map categorizes the state into distinct risk zones, aiding in hazard mitigation and land-use planning. The Frequency Ratio shows AUC of 0.760 and the Shannon Entropy shows the AUC of 0.724. Model validation through AUC- ROC analysis and success prediction rate curves confirm the reliability and predictive accuracy of the applied techniques. This comparative study helps in underscores the importance of combining statistical models with geospatial tools for disaster risk management and infrastructure planning in Sikkim. These maps can be further utilized by the government in mitigation planning.

2026-03-01T00:00:00Z SINGH, BALRAJ Minocha, Vijay K. (SUPERVISOR) http://dspace.dtu.ac.in:8080/jspui/handle/repository/23394 2026-10-16T06:35:33Z 2026-10-01T00:00:00Z

Title: ANALYSIS AND OPTIMIZATION OF SCOUR DEPTH AROUND T-SHAPE SPUR DIKES Authors: SINGH, BALRAJ; Minocha, Vijay K. (SUPERVISOR) Abstract: Alluvial rivers are prone to bank erosion and channel instability, a problem that worsens during high flows and floods, when rising velocity and bed shear stress can rapidly erode banks, shift the channel, and damage surrounding land and infrastructure. Spur dikes are a common river-training response to this, built to redirect the flow, protect the bank, and encourage sediment deposition. Their own use, however, brings a problem of its own: local scour at the structure, which if left unchecked can undermine the foundation and cause the spur dike to fail. Designing safer and more effective spur dikes therefore depends on a proper understanding of the flow and the scour it produces around them. This thesis examines how scour around a spur dike is shaped by its structural configuration, with a particular focus on improving traditional designs to reduce scour and give better bank protection. Its main contribution is a new configuration, the curved T-shape spur dike, obtained by curving the flange of the conventional T-shape, with the aim of spreading the flow around the structure more evenly, weakening the vortices it generates, and so reducing local scour at the head. These objectives were pursued through an experimental programme in a laboratory flume under clear-water conditions, comparing scour around the conventional I- and T-shape spur dikes with the proposed curved T-shape across a range of hydraulic conditions. The spacing between spur dikes placed in series was examined too, since it governs how scour develops between adjacent structures. Maximum scour depth, its development over time, and the resulting bed morphology were recorded throughout. Scour proved to be a progressive process, reaching equilibrium in around 8 hours in every run. The spur dike length (b) was fixed at 0.22 m, or 20% of the channel width, and of the variables examined, the flow condition U/Uc (the ratio of approach velocity to critical velocity) proved to be among the most influential: a higher flow intensity generates stronger vortices and moves more sediment, and so erodes the spur head more severely. Sixteen configurations of the curved T-shape spur dike were tested to find the best geometry, combining four flange v lengths, X = b/2 (0.11 m), 2b/3 (0.147 m), 5b/6 (0.183 m), and b (0.22 m), with four crest widths, Y = b/10 (0.022 m), 3b/20 (0.033 m), b/5 (0.044 m), and b/4 (0.055 m). X = b (0.22 m) with Y = b/5 (0.044 m) gave the best performance, with equilibrium scour depths of 0.142 m, 0.138 m, and 0.133 m at U/Uc of 0.95, 0.89, and 0.83, respectively. Spacing between spur dikes in series was tested at 2b, 2.5b, 3b, 3.5b, and 4b, and 3b proved optimum, cutting scour at the second spur dike by 50% relative to the first. The curved T-shape spur dike was then set against the conventional I- and T-shape spur dikes, and gave a lower scour depth under every hydraulic condition tested, an improvement traced to its curved flange, which redirects the approaching flow more smoothly toward the centre of the channel. CFD simulations reproduced this same behaviour, showing lower near-bed velocities, a weaker horseshoe vortex, and less flow separation around the curved T-shape than around the I- or T-shape spur dike. That the experiments and the simulations agree confirms the proposed design's ability to limit local scour and improve hydraulic performance, and points to the curved T shape spur dike as a way of strengthening bank protection and the long-term stability of alluvial channels. Machine learning was also used to estimate scour depth around spur dikes. A standalone Random Tree (RT) model overfit the data and generalised poorly, so it was hybridised with additive regression (AR), bagging (B), and random subspace (RSS) to give seven combinations – AR_RT, B_RT, RSS_RT, RSS_AR_RT, RSS_B_RT, B_AR_RT, and RSS_B_AR_RT – built on the key hydraulic and geometric parameters as input variables and evaluated for predictive performance. All seven performed well in training, and two carried this through to testing: B_AR_RT (R-Square = 0.9702, RMSE = 0.1305, NSE = 0.9692, and COC = 0.9845) and RSS_RT (R-Square = 0.9699, RMSE = 0.1321, NSE = 0.9685, and COC = 0.9841). Of these, only B_AR_RT held up as a reliable predictor of scour depth, with an R-Square of 0.9007, an RMSE of 0.1801, and a COC of 0.9491, suggesting that this hybrid approach can give dependable scour-depth estimates and stand in for extensive laboratory investigations at the preliminary design and analysis stage. Three regression equations, linear, non vi linear, and second-order polynomial, were also developed to predict scour depth around I- and curved T-shape spur dikes: the second-order polynomial gave the most accurate predictions for the I-shape, while for the T-shape, all three regression equations performed satisfactorily. Taken together, these results give a rounded picture of the hydraulic and scour processes at work around spur dikes, and show the curved T-shape spur dike to be an effective alternative to conventional configurations for reducing local scour.

2026-10-01T00:00:00Z