Sarangapani Palani

Network intrusion detection systems (NIDS) are critical for protecting modern network infrastructures from sophisticated cyber threats. This study presents a comprehensive comparative analysis of two advanced deep learning architectures: a pure Transformer-based model and a hybrid Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) model for multi-class intrusion detection. Using the preprocessed and balanced CSE-CIC-IDS 2018 dataset (Version 1, 2024), we evaluate both models across multiple performance metrics. The dataset underwent rigorous preprocessing including duplicate removal, feature engineering, and two-stage resampling (Random Undersampling and SMOTE) to address class imbalance. Our experimental results demonstrate that the Transformer model achieves 99.42% accuracy with a ROC-AUC of 99.58%, while the CNN-LSTM hybrid model achieves 99.18% accuracy with a ROC-AUC of 99.35%. The Transformer architecture excels in capturing long-range dependencies and parallel processing, while CNN-LSTM effectively combines spatial feature extraction with temporal sequence learning. Both models demonstrate superior performance in detecting minority attack classes compared to traditional approaches. This study provides insights into architectural trade-offs between self-attention mechanisms and recurrent-convolutional hybrids, offering practical guidance for implementing robust intrusion detection systems.

One of the practical components that can be found in aircraft engines, the Delaval nozzle, works to improve the aircraft's performance in terms of its stability and its capacity to manoeuvre at different speeds of operation. This research concentrates on the acoustic behaviors shown by this nozzle during the transition from low rate to sonic speed. The modification of the nozzle tip at various degrees of inclination on the transition velocity is the procedure used to minimize the amount of acoustic energy emitted. The acoustic behavior is analyzed based on the FW-H acoustics model utilized, and the optimization problem is solved using the practical eddy simulation approach in CFD. The acoustic Power, measured in dB, is estimated at various angles on the nozzle tip, and the performance of the Delaval nozzle is used to show how the redesigned nozzle tip affects the nozzle's acoustic behavior.

Nowadays, light duty diesel engines and lean burn petrol engines are getting much attention due to improved fuel economy. The present conventional catalytic converter controls effectively the levels of carbon monoxide and hydrocarbon, but it displays poor conversion in harmful oxides of nitrogen emission under lean exhaust condition. But the automobile pollution control regulatory bodies tighten the emission level every year. Zeolite-based catalysts have received a lot of focus recently because of their strong activity and comparatively broad temperature window. In the present work, zeolite-like material and ZSM-5 like material are synthesized by alkali fusion followed by hydrothermal treatment from coal flyash. Cupric Chloride (CuCl2) and Ferric Chloride (FeCl3) are used as transition metals. These metals are incorporated separately into the zeolites by conventional liquid phase ion-exchange method. Cu-ZSM5 (In house). Fe-ZSM5 (Inhouse), Cu-ZSM5 (commercial). FeZSM5 (commercial). Emission tests are conducted on CRDI engine. Initially the engine is run without catalytic converter in five different load conditions In all the cases, the concentration of CO, HC, O2, CO2 and NO, are measured by AVL Di-gas analyzer. It is observed that commercial catalytic converters at 16 kW load condition, the NO, conversion efficiency is 70%, 65% and 35%; the CO conversion efficiency is 95%, 80% and 82% and HC conversion efficiency is 93%, 79% & 80% respectively. For diesel engine, Cu-ZSMS (IH) like material is more effective in low temperature application.

Laser machining is a cutting-edge thermal energy-based industrial methodology that can be employed to replace the traditional mechanical machining practices in micro-domain applications. In this work, complex hexagonal shape profiles were cut on the surfaces of Nickel-based superalloy (Alloy 276) by fiber optics laser cutting system. Response surface methodology - Central Composite Design approaches were attempted for optimizing laser machining factors to obtain the better output responses. ANOVA was used to identify the significant machining factors which affect the cut quality characteristics. The desirability approach in multi-response optimization was deployed to achieve the minimal kerf width, low surface roughness, minimum heat affected zone while simultaneously increasing the material removal rate. To validate the predicted results obtained from the desirability index analysis, confirmation test was conducted based on a higher desirability index value. The result reveals that the error percentage is less than 5 for all the responses. The results of the confirmation test show that the proposed procedures are effective in improving laser machining parameters for obtaining the better responses.

Duplex stainless steel (DSS) is hard to machine by conventional methods due to its high strength combined with superior plasticity. The microdomain application of laser cutting is found in sheet metal industries owing to its high accuracy and precision. The current trend in laser technology is the use of fiber laser cutting (FLC) to produce complex shapes and structures. This work highlights the microhole cutting on 0.44 mm thick DSS by a 100 µm fiber laser beam. For optimizing the fiber laser parameters, the entropy with TOPSIS approach is employed. The response surface methodology is applied to find the regression model for FLC parameters, and ANOVA checks the adequacy of the model. The parameters considered for the current research work are the power of laser, frequency of pulse, speed of scanning, and duty cycle. The output responses are surface roughness, heat affected zone, overcut, and metal removal rate. The test results showed that the power of laser is the most prevailing factor that affects the quality of microholes. From the experimentations, it is found that experimental and predicted values are reasonably close to each other. The findings from confirmation experiment validates that the optimal condition produces better surface quality as compared to the initial trails.

The use of nickel aluminum bronze (NAB) is found in components that include landing gear bearings, cast valve bodies, and perforated orifice for pharmaceutical industries. Machining of the NAB alloy in the micro-domain is a great challenge because of its unique mechanical and chemical properties. This work investigates the suitability of electrode materials for anodic dissolution of the NAB alloy through the use of the electrochemical micromachining process. Three distinct electrode materials, namely, brass, copper, and tungsten carbide were chosen for understanding the micromachining characteristics. The nature of the dissolution of NAB alloy and the impact of all three electrodes has been studied by varying the process parameters. Micro-hole making experiments were conducted by response surface methodology (RSM) based central composite design (CCD). The RSM-CCD utilizes quantitative process parameters for choosing qualitative electrode materials based on radial overcut, surface roughness, and material removal rate. SEM pictures examined the morphologies of the micro-holes. All the three electrodes exhibited successful anodic dissolution of NAB alloy under various parametric conditions. NAB alloy machined with tungsten carbide electrode exhibited the least radial overcut and surface roughness, whereas the copper electrode removed the highest materials. The brass electrode exhibited intermediate performance between tungsten carbide and copper electrodes.