Papers by Keyword: Response Surface Method (RSM)

Abstract: The turning followed by burnishing is used like a combined process by using a combined tool having incorporated a turning tool and a burnishing system. This process is applied after different machining operations in a single hybrid process and can lead to the following advantages: the improvement of the part surfaces quality and to an important reduction of the costs and manufacturing times. The part surfaces quality obtained by applying the above mentioned process is influenced by a lot of factors, the most important among these being the working parameters. In the present paper there are presented the results concerning the determination of the optimal working parameters in order to obtain a high quality of the machined surfaces by applying the Response Surfaces method. The obtained results were verified by experimental tests and confirmed the efficacy of the method applied to optimize the results of the burnishing-turning combined process.

Abstract: The hardness of the coated surfaces by nickel-iron electroplating process is closely linked to the characteristics of deposited layer. These characteristics depend on various process variables, such as current density, temperature, pH and stirring. This work presents the modelling and the optimization of nickel-iron electroplating process variables to maximize the surface hardness. To study the combined effect of current density J (A/dm²), temperature T (°C) and pH were used a 2³ orthogonal central composite experimental design for experiments design and Response Surface Methodology for analysis of experimental results. The modelling was performed using the following intervals for process variables: 1.75 - 3.51 A/dm² for current density, 25-35 °C for temperature and 2.5-3.5 for pH. The empirical model was further used in the optimization process using the Gradient method. The optimum values of these electroplating process variables were found to be J = 2.23 A/dm², T = 30.80°C and pH = 2.81; in this point the surface hardness is 136.89 HV given by empirical model and 137.52 confirmed experimentally.

Abstract: Data-mining methods using hierarchical and non-hierarchical clustering are proposed, which could help manufacturing engineers determine guidelines for deciding end-milling conditions. We have constructed a novel system that uses clustering techniques and tool catalog data to support the determination of end-milling conditions for different types of recent difficult-to-cut materials. In the present report, we especially focus on the cutting speed to estimate the performance of this system. A comparison with the conditions recommended by famous tool makers in Japan, reveals that our proposed system can be used to determine the cutting speeds for various difficult-to-cut materials. That is, milling experiments using a square end mill under two sets of end-milling conditions (conditions derived from the end-milling condition decision support system and conditions suggested by expert engineers) for difficult-to-cut materials (austenite stainless steel; JIS SUS310) showed that the catalog mining method is effective for deriving guidelines for deciding end-milling conditions at the beginning of the manufacturing stage.

Abstract: In aim to gain the capability of providing variable thrust, the technology of an axial pintle inserted into the nozzle of the solid rocket motor had been used. As the pintle inserted into the nozzle, the loss of specific impulse will increase. In order to reduce the loss of fluid dynamics of the pintle nozzle, considering the interaction of the nozzle contour and the pintle contour, CFD method combined with the response surface method is used to optimize the contour of the pintle nozzle. The central composite design is used to introduce a design point; the Kriging algorithm is used to generate the response surface; and the Nonlinear Programming by Quadratic Lagrangian is used for optimization. After the optimization, the loss of fluid dynamics can be reduced significantly. To study the influence of the key parameters to the loss of specific impulse, the key parameters are optimized independently in this paper. It indicates that the main factor of the loss of specific impulse is the parameters of the nozzle. To reduce the computational consumption, the process of optimization has been improved. And the result shows that when optimizing a pintle nozzle, the nozzle part and the pintle part can be optimized separately. The method that bases on the response surface not only takes into account of the interactive effects of the shape parameters, but also works with less calculation. Additionally it maintains the high accuracy and reliability. It can be used to select the optimal shape parameters of the pintle nozzle quickly, which has certain engineering application value.

Abstract: The marine oil and gas transmission pipelines should be washed by phosphoric acid solution before they were used. The washing wastewater, containing high level of phosphorus, was treated by calcium chloride. The response surface method (RSM) based on Box-Behnken was adopted to consider the impact of n (Ca):n (P), pH and reaction temperature on total phosphorus (TP) removal rate. In this work, the Design-Expert 8.0.6 was employed to design and analyze the results. Then, a quadratic polynomial model was built with TP removal rate as the response value. It was shown that the n (Ca):n (P) and pH have a significant effect on TP removal rate. The interaction of pH and temperature is significant. The optimum conditions were X₁=3.40, X₂=10.97, X₃=38.12. In this condition, the predicted value of TP removal rate is 99.9242% while the average removal rate of verification is 99.9365%, which was shown that the predicted value was identical with experiment value.

Abstract: On the basis of ANSYS/LS-DYNA, the finite element model of incremented forming of conical part was constructed. The plastic deformation area of blank was increased and crack was delayed in multistage. Forming parts with the angle of 90°use multistage forming. Focused on the investigation of the influence of process parameters on the thinning rate and the thickness distribution then least-square response surface method was adopted to set up the approximate optimization model for thinning. The optimization on the approximate model was carried out using sequential quadratic programming method. The results show that the experimentally thinning rate is consistent with the theoretic result based on the homogeneous thickness distribution.

Abstract: Surface roughness is strongly affected by machining parameters. In the past few decades, many researchers have established the relationship between the surface roughness and machining parameters, but less attention has been paid to tool shape and geometry. In addition, the number of tool flutes was ignored, which affects in vibrations and machining system. Therefore, this study first-time includes the tool flutes in addition to cutting speed, depth of cut and feed rate as independent variables. Firstly, a set of machining experiments were conducted using AA6061 as a work piece material to provide original data. Response Surface Model (RSM) adopted to establish the relationship model of surface roughness and machining parameters using Minitab 16. Based on analysis of variance (ANOVA), the results show cutter flutes has higher significant followed by feed rate, depth of cut and cutting speed which has less significant. Finally, machining parameters were optimized to desired surface roughness, and optimization prediction error has limited values between-0.02 and 0.02μm.

Abstract: Powder forming process is used for fabrication of bulk ceramic components. Optimization of powder forming process parameters in the fabrication of alumina-zirconia composite used in orthopedic implants is done for desired physical property. In this research work, process parameters such as composition of zirconia, compaction pressure and sintering temperature were analyzed using Response Surface Method (RSM). The physical properties such as density, porosity and water absorption characteristics of the proposed composite were studied. To study the influence of the different process parameters over the physical properties of the fabricated composites materials, experimental runs were framed by using Box behnken method. Three factors and two levels were selected with a total of 17 runs and their consecutive tests were carried out. The validity of the model was checked and the significant parameters were identified using Analysis of variance (ANOVA). The results indicate that the sintering temperature is influencing predominantly the physical properties of composites compared to other process parameters.

Abstract: With the increasing of variables and orders of polynomial expansion, the undetermined coefficients of the response surface model have increased dramatically. The insignificant expansion terms should be eliminated because not all of the expansion terms have significant effect on the response. The improved and practical updating method based on response surface model by stepwise regression, which can effectively reduce the undetermined coefficients on the premise of guaranteeing the correcting accuracy, is proposed in this paper. The method is also applied to update the finite element model of a bridge with prestressed reinforced concrete rigid frame-continuous girders based on the in-situ dynamic test results. The results show that the updated finite element model can match the mechanical properties of the bridge well.

Abstract: Monte Carlo Simulation and Response Surface Method are two very powerful reliability analysis methods. Normally, in the reliability analysis of complex structures, the limit state function often can not be expressed in a closed-form. Usually, the codes for probabilistic analysis need to be combined with finite element models. ANSYS Probabilistic Design System (PDS) has provided a package to conduct probabilistic analysis automatically. This paper is going to compare the performance of these methods through an easy engineering problem in ANSYS. The results are going to be derived to show the feature of applying the corresponding reliability methods.