Papers by Keyword: Regression Model

Abstract: The V-bending process is one of the most important operations in sheet formation, which is influenced by a number of different factors. The phenomenon of springback is a well-known issue within the field of sheet metal forming, resulting in compromised dimensional accuracy and reduced operational efficiency. The objective of this research is to analyse and compare the springback properties exhibited by different types of sheet metal alloys (aluminium alloy 6061-T6, stainless steel 304, and low-carbon steel) with thicknesses (0.5, 1, 1.5, 2, 2.5 and 3) mm. The springback angle of each sample is subsequently evaluated quantitatively using a measuring machine of coordinate. A regression model is developed to predict the incidence of springback by using the variables of material properties and forming conditions. The results suggest that there is a significant variability in the springback phenomena, which is influenced by the particular alloy and thickness of the material. Specifically, the low-carbon steel has got the least springback, followed by SS 304, and Al 6061-T6. The thickness of the material has the greatest influence on the process (54%), followed by the type of material (37%). In contrast to the largest thickness of 3 mm, the maximum value of reflectivity is found in all three materials employed in this investigation, the smallest thickness of 0.5 mm provides the least amount of springback.

Abstract: Superalloys, referred to as nickel alloys, have several uses in engineering and are widely used in industries that are as diverse as chemical processing and food processing. The high thermal conductivity and high strength of these materials make them hard to remove material from with traditional processing techniques. The majority of modern techniques for machining harder materials are alternatives to older methods. The present study is focusing on Wire Electrical Discharge Machining (WEDM), a modern machining technique used for the processing of tougher materials. The aspiration of this work is to present a Taguchi-based Grey technique that can be used to optimize a number of different performance indicators. The connection between the input and output variables has been analyzed using a regression model. Taguchi's design approach has been applied to the design of trials, with the Pulse on/off time and the applied current serving as independent variables. For enhancing the multiple machining performance of nickel alloy during Wire Electrical Discharge Machining (WEDM), this experimental effort seeks to pinpoint the most effective variables. This is accomplished using the Taguchi-Grey method. The performance analysis offers producers with concrete proof of the efficiency of evolved systems, allowing them to make well-informed and effective choices.

Abstract: The purpose of this study is to investigate the friction stir weld quality of pulverized glass waste (PGW)-reinforced AA6061-T6 and develop a model predicting the hardness of the joint. Friction stir welding of PGW-reinforced AA6061-T6 was done within a process window. The process was optimized for the maximum joint hardness. Thereafter, the result of the hardness was used to develop a model using a novel statistical analytical technique. The addition of PGW enhanced the AA6061-T6 friction stir welded joint hardness. The maximum hardness (112 HV) of the PGW-reinforced joint, which was obtained at optimal setting of 900 rpm rotational speed, 40 mm/min traverse speed and 1^o tilt angle, is by a factor of 1.72 greater than the unreinforced joint and close to the hardness of the as-received AA6061-T6 (120 HV). The developed model can predict the hardness of the PGW-reinforced AA6061-T6 joint up to an accuracy of 89%. The model shows that the rotational speed, tilt angle and their interaction contributed significantly to the hardness of the PGW-reinforced AA6061-T6 friction stir welded joint. This model is suitable for determining the hardness property of particle-reinforced AA6061-T6 friction stir welded joint at varying processing parameters.

Abstract: A second-order regression polynomial model is offered, taking into account a pair components interaction for TCLE calculation of low-melting glasses in system MgSO₄-KPO₃-Na₂B₄O₇. The analysis of the obtained model is carried out. The results adequacy, obtained by modeling in comparison with the experimental data, is shown. It is established that the square pair interaction of the MgSO₄-KPO₃ and MgSO₄-Na₂B₄O₇ components, when modeling this system, can be neglected.

Abstract: A hydrocyclone is a device used widely in various industries, especially for separation of solids from liquids. Many factors affect the separation efficiency of a hydrocyclone. In this research, the main objectives were a study of the conical length that affected the separation efficiency and proposal of a regression model of Stk₅₀Eu for a hydrocyclone. First, research was performed on the separation efficiency using a 40-mm hydrocyclone. The effects of conical lengths of 200, 240 and 280 mm were investigated. The tested suspension was a mixture of silica and water. The silica particles have an average size of 9–10 μm at a solid concentration of 0.5% w/v. The feed-flow rate of 1 m³/hr was operated with the constant flow ratio of 0.1. From the experimental result, it was found that the shorter conical length obtained the higher separation efficiency. For a conical length of 200 mm, the cylindrical length of 60 mm and the vortex finder length of 40 mm showed the best separation efficiency, up to 84.06%. Second, a regression model of Stk₅₀Eu of the hydrocyclone was established. In this work, data obtained from a total of 75 experiments in the first part and from earlier research were used to form the relationship between the dimensions of the hydrocyclone and Stk₅₀Eu. The calculated Stk₅₀Eu can successfully be used to predict experimental Stk₅₀Eu.

Abstract: This study investigates cooling of water at night in Malaysian climate as renewable cooling medium source for radiant cooling purpose. An experiment with a 1.95 m² steel roof rig structure was constructed and night cooling cycle was conducted during the hot season and cold season of the year. Regression model was developed to predict water temperature after the night cooling process and the corresponding water and roof ratio was established. An annual simulation of a low income home model retrofitted with radiant cooling system charged by night cooled water as cooling medium shows that 99% of the time the thermal condition could meet ISO 7730 category C PMV between-0.7 and + 0.7 . For an outdoor ASHRAE design day condition, the peak indoor operative temperature of 37oC could be lowered to about 30oC with the use of radiant cooling system. The calculated energy saving for the home model was 85% or about 15% of the conventional air system operating cost.

Abstract: Production cost is dependent on the life of the Tool. Because of enormous heat generation during the material removal process, life of the tool decreases. Tool life will be enhanced by cryogenic treatment which minimises the temperature at tool tip interface. Taguchi technique was employed to get optimum number of experiments for turning white cast iron after the cryogenic treatment and before cryogenic treatment. The correlation between four main factors such as speed, feed, depth of cut, tool condition and responses such as surface roughness, tool tip temperature were analysed. Mathematical model was formulated for tool tip temperature, and surface roughness. The error for the mathematically formulated model was observed to be less than 5%.The present work indicates that cryogenically treated tool have better surface finish . From the anova analysis it is inferred that tool tip temperature and surface roughness substantially reduced while using cryogenically treated tool. It was observed that cutting forces was more influenced by cutting speed of the tool followed by tool condition. Hardness of the tool insert showed improvement because of coatings.

Abstract: The paper presents compare methods of approximation results of studies using regression analysis and neural networks. As a research facility used theoretical values of surface roughness based on theoretical values of surface roughness calculated as kinematics-geometric projection of the cutting edge on finish surface. Pointed to the limitations of the presented methods of research results approximation.

Abstract: Surface water film thickness is one of the main factors, which affect the vehicle safety on slippery roads. Water film depth is influenced by rainfall intensity, grades, cross slopes, drainage length and pavement texture. This paper reviews the research status and makes some comparative analysis of several pavement water film depth prediction models. An experimental validation has verified and calibrated the existing water film depth prediction models results. The experimental validation of the variable in the slope water flow model has been implemented by means of a small scale physical road model in a rainfall simulator, which is constructed in a laboratory. The results of comparative analysis have shown that in the existing water film depth prediction models, the regression models predict values are more closely than mathematical-physical models. Because under different experimental conditions, the regression model calibration parameters are different. In the case of specific road characteristics for prediction of water film thickness, the model parameters can be calibrated to further improve predicting accuracy.

Abstract: Based on the field survey data,statisticsing and analysising the influence factors of industry flows of science and technology human resources in Shaanxi Province by using multiple linear regression model and Logistic regression model.The results show that: enterprise promotion chance, industry development ability, family factors and change lifestyle these four factors promote or weaken the effect on the flow of talent directly;Age,working ability,professional degree are the constitute factors influencing science and technology talent flow;However, the influence of gender, education, welfare and residential moves is not significant for industry flow of science and technology human resources.