Knowledge Connect @ SASTRA
Not a member yet
201 research outputs found
Sort by
Farmers’ Hiring Decisions towards Farm Mechanization Services in the Cauvery Delta Zone of Tamil Nadu
No full textAgriculture is critical to any country\u27s economic development. It not only benefits the primary and secondary sectors but also helps the economy flourish. In addition to ensuring food security, the Indian agricultural sector provides work to huge populations, both directly and indirectly. Traditional farming practices are now overused, labour is no longer abundant, and the power demand is rising steadily. The rapid development of technology has given rise to new opportunities in agriculture (Balafoutis, Evert, & Fountas, 2020). Technology aids agricultural growth by allowing more effective utilization of inputs. It has the potential to make an immense contribution to agricultural development. A significant number of farmers are facing numerous difficulties, such as varied topography, small and fragmented farm holdings, and lack of investments and technology to follow subsistence agriculture and use conventional farm machinery. Agricultural productivity is affected due to the prevalence of impediments like land fragmentation, labour scarcity, increase in wages for farm labour, climate change, lack of infrastructure, credit facilities, etc., In addition to combating these impediments, a technologically driven concept called \u27Farm Mechanization\u27 is a crucial farming approach that helps to improve agricultural productivity.
This study used a few constructs from the Innovation Diffusion Theory, Technology Acceptance Model, and Theory of Planned Behaviour models, which affect farmers\u27 hiring decisions for mechanisation services. The constructs used for the Technological attributes include Relative Advantage, Perceived Convenience, Perceived Economic Cost Benefits, Perceived Usefulness, Trialability, and Observability. The Psychological factors have constructs: Attitude, Subjective Norms, Perceived Behavioural Control, and Values. The institutional factors have constructs Access to Informational sources, Service Providers, Access to credit through Financial Institutions, Government Support, Extension Services, and Environmental factors. The Socio-economic variables used for the study include Age, Education, annual income, farm experience and Land Holding.
The researcher has collected the data from the farmers of the Cauvery Delta zone in Tamil Nadu, which is considered as the population for the study, by applying the multistage cluster sampling method. The structured questionnaires were distributed to 1010 farmers across the Cauvery Delta Zone consisting of Thanjavur, Tiruvarur, Nagapattinam, Mayiladuthurai, Tiruchirapalli, and Ariyalur districts. Kulithalai Taluk of Karur district, Aranthangi Taluk of Pudukottai district, and Chidambaram and Kattumannarkoil Taluks of Cuddalore District. A total of 854 questionnaires were received, with a response rate of 84.55%. 44 questionnaires were found with some missing information after the cleaning process. Finally, 810 questionnaires were found to be fit and included in the analysis.
Out of eighteen factors of hire purchase decision, The topmost five factors are environmental factors (3.827), subjective norms (3.703), followed by farmers\u27 intention on hire purchase(3.697), farmers\u27 actual hiring (3.681) and perceived behaviour control (3.656). Whereas the least inflowing factors are Values (3.543), Perceived convenience (3.511), Economic Cost benefits (3.458), Extension Services (3.455) and Service Providers (3.362).
All sub-factors of technological and institutional factors show a high positive correlation with farmers\u27 Intention to hire. The respective Pearson correlation values for the Psychological factors such as ‘Attitude’, ‘Subjective Norms’ and ‘Values’ denote that they have a high positive correlation with ‘Intention’ and ‘Perceived Behavioral Control’ has a moderate positive correlation with ‘Intention’.
Environmental factors, Access to Informational sources, Perceived Behavioural Control, Perceived Economic Cost Benefits, Observability, Relative Advantage and Trialability have significant influence on Intention. Further, Intention is found to have a significant influence on Actual hiring decisions. The R-Square value of 0.708 represents that 70.8% of the variation in the \u27Intention\u27 can be explained and caused by technological attributes, psychological factors, and institutional factors. Likewise, the R-Square value of 0.635 of the second-order construct represents that 63.5% of the variation in the Actual hiring decisions can be explained and caused by Intention.
There is a significant influence of Technological attributes, Psychological factors and Institutional factors on the dependent variable \u27Intention\u27, and there is a significant influence of Intention on Actual hiring decisions, respectively. The R-Square value of 0.646 represents that 64.6% of the variation in the \u27Intention\u27 can be explained and caused by variables such as Technological attributes, Psychological factors and Institutional factors. Likewise, the R-Square value of 0.634 of the second-order construct represents that 63.4% of the variation in the Actual hiring decisions can be explained and caused by Intention.
In terms of the moderating effect on socioeconomic factors towards actual hiring: annual income and landholding have a significant effect. In contrast, the other factors, age, education, and farming experience, have no significant moderating effect on actual hiring decisions.
The study confirms that different socio-economic profiles have different effects on these three factors. The correlation and structural equation analyses also confirmed that there is a constructive effect between these three factors: Technology, Psychological, and Institutional and the farmer\u27s Intention and actual decision to hire machinery
Hormetic Modulation of on mTOR Mitochondria Cross Talk for Mitigation of Cellular Aging
No full textCellular aging is a gradual process involving structural and functional deterioration of cells. Mitochondria, the biological clock in cells, are known to play a crucial role in the progression of cell aging. According to the mitochondrial free radical theory of aging, oxidative phosphorylation in the mitochondria produces free radicals, such as, reactive oxygen species (ROS), which lead to oxidative stress in cells. This, in turn, results in mitochondrial dysregulation, and eventually causes degeneration of cellular components, thereby propagating cell aging. However, it is to be underscored that aging, at its core, is a multifaceted phenomenon that involves a number of complex cross-talks of diverse aspects within a cell.
While it has been known that improved mitochondrial function is instrumental in mitigating cell aging, some studies have correlated a reduced rate of aging with partial/incomplete inhibition of mTOR activity. Studies conducted in the Intervention Testing Program of the National Institute of Aging (NIA-ITP) showed an increase in longevity in animals on treatment with mild doses of rapamycin (RAP), which is a well-known inhibitor of the multifaceted cellular regulator, mTOR. Intriguingly, RAP is also known to be extensively toxic towards cells. There is, thus, a clear dichotomy in its functions — it is cytotoxic at moderate/high doses, while it can restrain aging and extend lifespan at low doses. Such biphasic effect of drugs is known as hormesis.
Earlier studies have suggested that the hormetic (low) doses of RAP might cause partial/incomplete inhibition of mTOR, which may be attributed for the reported life-prolonging effect. However, understanding is still limited on the actual interplay of mitochondria and mTOR in the precincts of aging. The actual modus operandi of how such partial mTOR inhibition might modulate the mTOR-mitochondria cross-talk remained to be deciphered in the context of cellular aging. It is, therefore, prudent to investigate the mechanisms and impacts on cell aging by modulating facets of mTOR-mitochondria cross-talk via hormetic effects of RAP, and its analogs (rapalogs), such as temsirolimus, everolimus, and ridaforolimus (RFL), a non-prodrug rapalog.
Although prior reports have stated the role of ‘mild’ or incomplete inhibition of mTOR in slowing down of the aging process, till date no study has reported ‘an mTOR-mitochondria cross-talk’ that governs cellular aging. Moreover, it was also unknown how ‘mild’ mTOR inhibition affects the mitochondria-governed aging. We hypothesized that ‘mild’ degrees of mTOR inhibition (incomplete inhibition) by the hormetic function of RAP/RFL can cause mitigation of oxidative stress-governed injury to the mitochondrial paraphernalia, thereby alleviating degenerative conditions in cells and augmenting cellular longevity. We aimed to investigate an ‘mTOR- mitochondria cross-talk’ by administration of very low doses of RAP and RFL. No prior report was available on the molecular roleplay of RAP and RFL in the cross-talk between mTOR and mitochondria in the aging process. Furthermore, no literature showed the specialized hormetic function (at very low doses) of RFL and RAP on the molecular modulation of the mTOR-mitochondria functional axis that we have hypothesized.
In the undertaken study, we investigated the hormetic nature of RAP and RFL, in modulating the mTOR-mitochondria cross-talk to manifest an anti-aging outcome in WRL-68 cells. Cells treated with low doses of RAP/RFL were explored to comprehend how the hormetic function of the drugs affected the aging-associated parameters, such as, mitochondrial metabolism and biogenesis, mitophagy, oxidative damage to mtDNA/mtRNA and mitochondrial proteins, viability, cell death and senescence, mitochondrial density and Δψm. In conclusion, we report that low doses of RAP and RFL can hormetically amend the mTOR-mitochondria cross-talk, and can consequently promote anti-aging outcome in cells
YieldNet: Intelligent Fruit Yield Estimation for Selected Orchards using Deep Learning Based Semantic Segmentation
No full textAgriculture contributes more resources for developing sustainable economic growth of the nation. Precision agriculture employs advanced techniques (machine learning and deep learning) for developing the intelligent systems of various agricultural applications. Among various agricultural tasks, yield estimation of crops plays a vital role in decision-making such as harvesting, marketing, cultivation practices, etc. Traditionally yield estimation is performed manually which has major drawbacks i.e., needs experts opinion, time-consuming and it is a challenging task for big orchards. To overcome these issues, an intelligent yield estimation model using neural network-based systems is required.
Some of the literature works have been explored for fruit yield estimation using intelligent techniques namely, deep learning-based semantic segmentation architectures. But, the analysis of a customized counting model which gives better localization for mapping the fruit yield including challenging situations like partially occluded and overlapped conditions is yet to be explored. Hence, the objective of the work is to develop an intelligent fruit yield estimation for selected orchards (tomato and mango) using deep learning-based semantic segmentation architectures.
In the first phase of work, tomato yield estimation was performed using three deep learning-based semantic segmentation architectures such as U-Net, SegNet with VGG16 and SegNet with VGG19. Training was done on the dataset of 672 tomato images and testing was done on the real-time field data of 65 tomato images. The test results revealed the highest precision, recall and F1-score values of 89.7%, 72.55% and 80.22%, respectively for the SegNet with VGG19 architecture among the three architectures compared.
Finally, the segmented fruits were counted using the contour detection method. Then, the final yield in kilograms was estimated for the chosen tomato field. The error percentage between actual and predicted weight is 4.8%. A user-friendly graphical user interface was developed for estimating the tomato yield. However, as VGG19 is used as the backbone network for SegNet, execution time and memory consumption are the hurdles for real-time implementation.
To overcome these issues, suitable advanced deep learning-based semantic segmentation architecture can be employed. Hence, in the second phase of the work, MangoYieldNet for intelligent counting of mangoes using DeepLabv3+ was proposed which extracts the features from the mango images by employing atrous spatial pyramid pooling and an effective decoder module for better localization of mangoes. The mango images were taken during the daytime on both sides of the trees.
After the sampling process, 556 images (from 278 trees) were captured and annotated. Using image augmentation techniques (reflection, rotation and translation) a dataset of 1152 images was developed; it was split in an 80:20 ratio for training and validation, respectively. The training was initiated using random weight parameters and the hyper-parameters were optimized using stochastic gradient descent by minimizing the errors at the epoch of 50. For testing the trained architecture, 30 new images were captured from the mango orchard. The test dataset revealed the highest mean accuracy of 96.5% and a mean intersection over union of 95.97% for the architecture of DeepLabv3+ with ResNet18 and better object localization among other architectures (i.e., DeepLabv3+ with MobileNetv2, DeepLabv3+ with Xception, U-Net and SegNet with VGG19) compared.
Finally, the segmented fruits were counted using the circle Hough transform method. The mango count was compared with the manual count obtained from the farmers and provided the regression coefficient of 0.98. Then, the final yield in kilograms was estimated for the sampled trees in the orchard. The error percentage between actual and predicted weight is 4.99%. A user-friendly graphical user interface was also developed for estimating the mango yield. The future scope includes extending the work for estimating the yield for other fruits and deploying the model to robotic harvesting systems
Experimental Investigation on Mechanical Properties of PLA Reinforced by Bamboo Fiber/Montmorillonite Clay Hybrid Composites and Analysis of Machining Characteristics
No full textThe study presents a comprehensive investigation into the development, characterization, and machining analysis of novel bio-based polymer hybrid composites comprising Polylactic acid (PLA), bamboo powders (BP), and montmorillonite clay particles (MMT). Initially, the hybrid composites were fabricated using a solvent-free stir casting process, optimizing the weight percentages of BP and MMT to enhance mechanical and thermal properties.
Mechanical tests, including tensile, flexural, and impact tests, along with thermal analysis using thermogravimetric analysis and differential scanning calorimetry, were conducted to evaluate the performance of the composites. In the subsequent phases, advanced machining techniques were employed to analyze the machining characteristics of the optimized hybrid composite. Abrasive water jet machining was utilized to systematically assess surface roughness, kerf angle, and material removal rate.
A Box-Behnken design of experiments, combined with Response Surface Methodology (RSM), Analysis of Variance (ANOVA) and Multi- Objective Particle Swarm Optimization (MOPSO), allowed for the identification of the best machining parameters, including traverse rate, abrasive feed rate, and standoff distance, to achieve excellent machining performance.
In the final phase, laser machining was investigated as a precise material cutting technique for the PLA/BP/clay hybrid composite. Machinability evaluation focused on key parameters including laser power, scan speed, and gas pressure, with ANOVA utilized to understand their impact on quality characteristics. A model that combines the RSM, ANOVA, and MOPSO algorithms was employed to predict and improve the quality parameters of laser cutting
An Efficient Regression Testing Suite Optimization System With ISO Quality Factors
Get PDFRegression testing is a black-box testing technique. It is utilized to validate an alteration in code in the software to ensure whether it has affected the present performance of the product. It has also been used to assess the adjusted variants of the product. Moreover, software testing is the most efficient process in Software Development Life Cycle (SLDC).
The study introduces Green cloud computing, incorporating computer resources such as foundations, PCs, application administrations, and information stockpiling. Notably, the research imbibes reliability, dependability, and maintainability as quality meters in the validation process. The goal of the proposed system is to implement Software Regression Testing in less time and at a low cost to transform the quality. The objective of the proposed study is to improve the optimization process, which employs the Cuckoo search algorithm, hybrid PSO, IPSO Algorithm, and TW-GA method.
The study examines regression testing to decrease the time and cost. Considering the regression testing, the proposed study intends to evaluate and discover prevailing secondary studies on software testing quality. It has been proved that the proposed study will enhance fault detection by identifying test cases from the prohibited detection capabilities
Design and Development of a Clinical Decision Support System for the Diagnosis of Parkinson’s Disease using Artificial Intelligence
Get PDFParkinson’s disease (PD) is a degenerative neurological condition marked by motor symptoms like tremors, bradykinesia, and stiffness. It is observed that early and precise diagnosis of this disease is crucial, as that will have a significant impact on effective disease management and intervention. This thesis explores the technical feasibility of applying AI techniques to recognize patterns from spiral and wave drawings, which are usually a unique signature type for PD patients.
The focus of the work is to diagnose the disease through novel deep transfer learning techniques, to diagnose the severity of the disease, and also to develop effective model interpretability techniques to improve the reliability of the trained models. The study gathers spiral and wave drawing data from PD patients and healthy controls from standard benchmarked data sets, followed by applying standard preprocessing and training techniques to extract relevant characteristic features.
Advanced transfer learning techniques, such as fine-tuning and pre-trained convolutional neural network (CNN) models, are applied to identify PD-related patterns automatically. Model interpretability techniques are then employed to understand learned representations and discover classification aspects, enhancing the identification of significant motor and non-motor symptoms linked to PD-related drawing anomalies. The proposed approach showcases a comprehensive hybrid deep transfer learning model incorporating explainable AI (XAI), which has been created for the early detection of Parkinson’s Disease (PD), achieving a classification accuracy of 98.45%.
Furthermore, this thesis evaluates the effectiveness of gait signals in PD diagnosis using recurrent neural network (RNN), long short-term memory (LSTM), and gated recurrent unit (GRU) architectures. Data collected from PD patients and healthy individuals are pre-processed, and relevant features are extracted for model training. The models are trained to identify PD-related patterns autonomously. Each model is evaluated based on accuracy, sensitivity, and specificity. Results indicate that RNN, LSTM, and GRU architectures effectively diagnose PD using gait signals, accurately differentiating between PD and non-PD instances. The GRU model achieves a higher classification accuracy of 98.20% in accurately classifying PD. The experimental results prove the superiority of the suggested gait analysis method in accurately predicting the severity of Parkinson’s disease. The accuracy rates achieved were 96.34% on the H&Y scale and 97.14% on the UPDRS scale.
Further, to investigate the impact of AI on multimodal data, the research focused on using PET images from PD patients and healthy controls to diagnose PD. Model interpretability techniques, such as LIME and SHAP, help identify regions of interest and factors influencing classification decisions, aiding in understanding the underlying pathophysiology of PD and developing biomarkers for disease progression. This will ensure that we understand how models come to their conclusions. Most importantly, the explainable AI techniques known as LIME highlight the most critical regions contributing to the PD classification, yielding a higher accuracy rate of 94.4% using Alex Net for the proposed PD classification method
Imperial Anarchy, Cultural Exchange and Cultural Sabotage in Amitav Ghosh’s Ibis Trilogy
No full textThe present study aims at analysing three sequential novels written by Amitav Ghosh. The novels are set in the backdrop of colonized India, so the research begins by examining the imperial anarchy that leads to cultural exchange and cultural sabotage. Amitav Ghosh is an Indian born writer, famous for both his fictions and non-fictions.
Most of his works are based on colonization and its aftermath. The major focus of the novels selected for the research is also colonization especially the opium production in India. The author portrays the opium production and its evil effects on people and environment. The three sequential novels include Sea of Poppies, River of Smoke and Flood of Fire.
The novels start with the opium production in India and it ends with the opium war in China. Opium forms the predominant theme of all three novels. Many research articles published on Ibis Trilogy are collected and it is evident that most of the articles deal with opium’s influence on flora and fauna, condition of Diasporas and language of Ghosh. No work of meritable length has been published on Cultural exchange that occurs in Trilogy. Hence the research tries to fill the gap.
Chapter 1 : Deals with the introduction. It includes introduction to Indian English Literature, author, and the sequential novel Ibis Trilogy. A short introduction on the works of Amitav Ghosh is also included. Literature survey has been done on various aspects related to the research that includes Ibis Trilogy and other novels of Ghosh, British Imperialism, Culture and opium. Based on the survey, the research gap is identified.
Chapter 2 : Deals with history of colonized India and opium production during 18th century. This helps in better understanding the novel. The chapter also explains the theory of cultural exchange. So the basic elements for understanding British imperialism, cultural exchange and cultural sabotage are included in this chapter.
Chapter 3 : Includes the views of historians and critics on the British rule, and Ghosh’s opinion on colonisation. The brutal rule of Britain as narrated by the novel, is also included in this chapter.
Chapter 4 : Talks about the cultural exchange that takes place in the novel. It includes the cultural exchange between Britain and India in colonized India and Britain, China and India during the time of opium war. The reason behind the cultural exchange and the results and impacts of cultural exchange are included.
Chapter 5 : Focuses on the sabotage of culture due to the greed for money. An insight on the true cultural values is included. The way the cultural essence is spoiled is analysed.
Chapter 6 : The conclusion traces the work beyond its writing. It ends with a note of optimism though the climax of the novel seems negative
Development of an Efficient Multi-Objective Approach for Secure Live Virtual Machine Migration
No full textCloud computing offers organizations flexibility and cost-efficiency through pay-asyou- go services, allowing them to scale resources according to their needs and reduce expenditures. Cloud as a Service (CaaS) offloads IT management complexities, while Cloud Data Center (CDC) provides infrastructure for on-demand, scalable, and flexible services over the Internet. Virtualization improves operational efficiency by providing simultaneous access to multiple virtual machines, while Live Virtual Machine Migration enhances agility, resilience, resource allocation, and fault tolerance.
However, achieving effective VMM requires forecasting cloud resource utilization, selecting the right target host, and ensuring security. Live VM migration is inevitable for optimizing CDC resource utilization. A Hypergraph-based Convolutional Deep Bi-Directional Long Short-Term Memory (CDB-LSTM) model is developed to predict resource usage. The model uses the Helly characteristic of Hypergraph to extract informative samples and the Savitzky–Golay filter to eliminate noises. The prediction approach uses the correlation coefficient measure to select the appropriate source VM and potential destination servers for migration. The model enhances resource usage prediction accuracy, reduces migrations, and preserves minimal computational costs during VMM.
However, it fails to determine the most secure destination with a lower total migration time. Selecting the optimal destination for live VM migration is a complex task that requires maximizing resource efficiency, minimizing energy consumption, and ensuring security. A framework for CDC uses the Network-aware Dynamic multi-objective Cuckoo Search algorithm (NDCS), a bio-inspired metaheuristic optimization algorithm, and an adaptive step-size approach. The hybrid movement strategy and fitness function resulted in the best-suited physical machine with the shortest migration time and lowest risk score. The proposed approach ensures security with fast convergence compared to existing methods. The system\u27s efficacy was assessed using the Google cluster dataset, showing it outperforms existing methods regarding reduced makespan time, energy consumption, and total migration time. However, the system still needs to ensure the security of the VM to be migrated. The security of VMs during migration is essential to prevent Service Level Agreement violations. A deep learning-based DDoS classification system is developed to detect if the migrating VM is free from DDoS attacks. Cryptographic methods are employed to protect the migrating VM from vulnerabilities.
An Improved Sparrow Search-based Deep Neural Network (ISSA-DNN) is used for DDoS attack classification. Advanced Encryption Standard-Elliptic Curve Cryptography (AES-ECC) is implemented to ensure VM image security. Preprocessing tasks like removing duplicates, adopting a Random Forest for feature selection, and normalizing the CIC-DDoS dataset were performed to improve the model. The DNN classifier outperformed existing methods, and encrypting VMs for migration is a proactive security strategy that protects sensitive data and ensures compliance with regulations. Compared to state-of-the-art techniques, the developed method reduces encryption and decryption time and increases throughput through effective key generation schemes
Impaired Speech Recognition of Neurological Disorder Persons Using Machine Learning and Deep Learning Techniques
No full textSpeech Assistive Tools have emerged in recent years to support individuals with cognitive and neurological disorders in the field of assistive technology. People affected by neurological disorders such as autism, stroke, cerebral palsy, dysarthria, Parkinson’s disease, and brain injury often find it difficult to articulate desired sounds, resulting in impaired speech. As the population of impaired speakers continues to increase every year, there is a strong need to develop intelligent speech recognition systems for affected individuals. The primary objective of this research is to develop an Impaired Speech Recognition (ISR) system for the Tamil language. Word Recognition Accuracy (WRA) is used as the performance metric, and a new dataset called the Impaired Speech Corpus in Tamil is created using speech samples collected from individuals with varying neurological disorders and intelligibility levels.
The proposed ISR system incorporates a Deep Neural Network–Hidden Markov Model (DNN-HMM) framework trained using the Lattice Free Maximum Mutual Information (LF-MMI) approach for effective recognition of impaired Tamil speech. Training and testing samples are collected from speakers with high, medium, low, and very low intelligibility levels. The recognition performance is evaluated and compared with baseline approaches using two datasets: a 20-word acoustically similar word dataset and a 50-word Impaired Speech Corpus in Tamil.
To address noisy, incomplete, and severely degraded impaired speech samples, an Enhancement Generative Adversarial Network (EGAN) is proposed for waveform enhancement. This approach improves the quality of impaired speech utterances and leads to better recognition performance on both the Tamil impaired speech datasets and the Universal Access benchmark database. The enhanced speech signals contribute to improved robustness and accuracy in impaired speech recognition.
Learning compact and efficient representations for disordered speech is challenging due to limited availability of impaired speech data. To overcome this issue, a novel sequence-to-vector representation based on HMM state sequences (HMM-SS) is proposed. This compact representation performs effectively with small datasets and is evaluated using four datasets: 50 words from TORGO, 100 common words from UA-SPEECH, 50 help-seeking words, and 100 common words from the Tamil impaired speech corpus. The proposed approach consistently outperforms baseline HMM, DNN-HMM, and state-of-the-art methods.
Finally, self-supervised and spectrogram-based approaches are explored to further improve impaired speech recognition. A Self-Supervised Learning (SSL) based Wav2Word framework using the wav2vec 2.0 encoder is proposed and evaluated on Tamil and English impaired speech datasets, achieving superior performance over conventional methods. In addition, a Denoising Convolutional Autoencoder (DCAE) is introduced to enhance spectrogram representations prior to CNN-based recognition. The proposed DCAE approach achieves significant performance improvements, with a maximum Word Recognition Accuracy of 96.07% on the Impaired Speech Corpus in Tamil, demonstrating its effectiveness for rehabilitation-oriented assistive technologies
Experimental Investigation and Evaluation of Mechanical Properties and Microstructural Characterization on Chromium Molybdenum Steel Processed through LASER based Welding
Get PDFCreep Strength Enhanced Ferritic (CSEF) martensitic modified 9Cr-1Mo (P91) steel is extensively employed in fast breeder nuclear reactors and thermal power plants owing to its superior resistance to thermal fatigue and oxidation corrosion compared with other low-alloy steels. Despite its excellent high-temperature performance, the weldability of this steel remains a critical concern.
Conventional arc welding processes generate excessive heat input and complex thermal cycles, resulting in heterogeneous microstructures, retention of δ-ferrite, and subsequent degradation in service life. In contrast, high energy density beam welding techniques such as laser and electron beam welding offer precise thermal control and minimal heat-affected zones (HAZ). However, research on the laser welding of P91 steel remains limited. The present investigation aims to comprehensively evaluate and optimize laser welding and post-weld heat treatment (PWHT) parameters for P91 steel using Response Surface Methodology (RSM).
The study was conducted in three distinct phases. Phase I focused on optimizing laser power (LP), welding speed (v), and focal position (FP) to achieve desirable weld bead geometry under argon and carbon dioxide shielding atmospheres. Analysis of variance (ANOVA) identified welding speed as the most influential parameter affecting penetration depth, bead width, and HAZ width. Optimum welding conditions yielded maximum penetration depths of 5.5 mm (argon) and 6.6 mm (CO₂) at low heat inputs of 0.077–0.094 kJ/mm, corresponding to minimal HAZ widths of 0.36 mm and 0.20 mm, respectively.
Phase II investigated the influence of tempering temperature, holding time, and heating rate during PWHT on the hardness and microstructure of the fusion zone. Hardness decreased with increasing temperature and time, while FESEM–EDS analyses confirmed the formation of a uniform tempered microstructure characterized by the redistribution of M₂₃C₆ and MX-type precipitates and the reduction of internal strain.
Phase III evaluated the mechanical performance of as-welded and PWHT joints. The PWHT specimens exhibited significant improvements in impact toughness (61% for argon and 41% for CO₂ shielding), alongside acceptable tensile strength and bend ductility. Fractography revealed ductile fracture features, and residual stress analysis showed a shift from compressive to tensile nature after PWHT.
Overall, the optimized fiber laser welding and PWHT conditions produced deep-penetration, low-heat-input welds with narrow HAZ and homogenized microstructure, thereby enhancing mechanical performance and demonstrating their suitability for high-temperature service in power plant applications