Response surface methodology for advanced manufacturing technology optimization: theoretical fundamentals, practical guidelines, and survey literature review

Oliveira, Lucas Guedes de; Paiva, Anderson Paulo de; Balestrassi, Pedro Paulo; Ferreira, João Roberto; Costa, Sebastião Carlos da; Campos, Paulo Henrique da Silva

doi:10.1007/s00170-019-03809-9

Cited by 80 publications

(50 citation statements)

References 130 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The higher the response surface slope, the more significant the interaction, and vice versa. The outline of the contour map can visually reflect the degree of interaction between each pair of factors, with not significant interactions represented by a circular shape and significant interactions by an oval shape [19]. There exists an obvious interaction for every two factors out of three, ranging from significant to un-conspicuous, e.g., AB > BC > AC.…”

Section: Regression Analysis Of Response Surface Test Resultsmentioning

confidence: 99%

A Study of the Electrodeposition of Gold Process in Iodine Leaching Solution

Meng

et al. 2019

Metals

View full text Add to dashboard Cite

This paper mainly discusses electrodeposition of gold from iodine leaching solution through single-factor testing and explores the influence of gold concentration in catholyte, mass fraction of iodine in anolyte, anolyte n(I2):n(I−), cell voltage, and electrolysis time on the percentage of gold deposition and coulombic efficiency. Moreover, a response surface methodology was adopted to study interactions among variables and influence on the percentage of gold deposition, with mass fraction of iodine in anolyte, anolyte n(I2):n(I−), and cell voltage as variable factors and percentage of gold deposition as the response value. The electrodeposition process was fitted via pseudo first-order kinetics and pseudo second-order kinetics. Finally, the free surface morphology of gold deposited on the cathode plate was observed by scanning electron microscope. Given the results, a principal effect relationship can be concluded between mass fraction of iodine in anolyte, anolyte n(I2):n(I−), cell voltage, and percentage of gold deposition, with cell voltage > anolyte n(I2):n(I−) > mass fraction of iodine in anolyte and a second-order regression equation obtained with percentage of gold deposition as the response value. The optimized process conditions were gold concentration in catholyte 20 mg/L, mass fraction of iodine in anolyte 0.59%, anolyte n(I2):n(I−) is 1:7.5, cell voltage 12.9 V, and electrolysis time 2 h. The average percentage of gold deposition of three confirmatory experiments was 96.43%, a figure very close to the predicted value of the model 97.76%, which proves that the quadratic polynomial model obtained by response surface methodology optimization is feasible and that the electrodeposition of gold process conforms to a pseudo second-order kinetic model. Au can be attached well to the cathode plate and the deposition layer is formed by granular grain accumulation.

show abstract

Section: Regression Analysis Of Response Surface Test Resultsmentioning

confidence: 99%

A Study of the Electrodeposition of Gold Process in Iodine Leaching Solution

Meng

et al. 2019

Metals

View full text Add to dashboard Cite

show abstract

“…Before performing the optimal design of heating system for a RTCM mold with the developed internal induction heating, the quantitative relationship between the objective functions and design variables should be established firstly. For this reason, the RSM was adopted in this work for its simplicity and solid mathematical basis [26]. There are many types of RSM can be used, but the second-order polynomial RSM is the most commonly-used one in various applications, which can be expressed as:…”

Section: Establishment and Validation Of Response Surface Modelsmentioning

confidence: 99%

Multi-objective optimization of heating system for rapid thermal cycling molding mold with internal induction heating

Xiao

Kahve

2021

SN Appl. Sci.

View full text Add to dashboard Cite

A rapid thermal cycling molding (RTCM) with novel internal induction heating mode is proposed in this work. The induction coils are directly inserted in the corresponding mounting holes of mold with an annular gap in between. During mold heating, eddy current losses confined at the walls of the mounting holes act as thermal sources to rapidly heat the mold cavity surface. Water passed through the annular gaps can be utilized to cool the mold in the cooling stage. Moreover, a design framework of the internal induction heating system in the RTCM mold is also developed. Firstly, a unit cell model of the mold was established to evaluate mold thermal response via numerical simulations, in which the effect of frequency and magnitude of coil current, the layout of induction coils and the annular gap size were examined. Then, a hybrid multi-objective optimization method was applied to optimize the induction heating system for the unit cell model. Finally, based on the obtained optimal parameters, a novel design strategy was adopted to conformally arrange the induction coils for the industrial RTCM molds. The blow mold of automotive spoiler was taken as an example to validate the effectiveness of the proposed approach. The results show that the present approach cannot only improve the mold thermal response performance, but also facilitate the mold heating system design process. This work may provide an effective method to realize RTCM of industrial plastics parts with free-form shape.

show abstract

“…non-linear systems). Especially with the rapid development of information technology in the recent few decades, the utilization of RMS has been spreading to cover many other fields such as civil [14], advanced manufacturing [15,16], and biomedical engineering [17,18] and agricultural and food science [19,20]. Experimental data has become much easier to collect, process and cache, parallel to which is the emergence of machine learning.…”

Section: Introductionmentioning

confidence: 99%

Introducing Machine Learning Models to Response Surface Methodologies

Zhang¹,

Wu²

2021

Response Surface Methodology in Engineering Science

View full text Add to dashboard Cite

Traditional response surface methodology (RSM) has utilized the ordinary least squared (OLS) technique to numerically estimate the coefficients for multiple influence factors to achieve the values of the responsive factor while considering the intersection and quadratic terms of the influencers if any. With the emergence and popularization of machine learning (ML), more competitive methods has been developed which can be adopted to complement or replace the tradition RSM method, i.e. the OLS with or without the polynomial terms. In this chapter, several commonly used regression models in the ML including the improved linear models (the least absolute shrinkage and selection operator model and the generalized linear model), the decision trees family (decision trees, random forests and gradient boosting trees), the model of the neural nets, (the multi-layer perceptrons) and the support vector machine will be introduced. Those ML models will provide a more flexible way to estimate the response surface function that is difficult to be represented by a polynomial as deployed in the traditional RSM. The advantage of the ML models in predicting precise response factor values is then demonstrated by implementation on an engineering case study. The case study has shown that the various choices of the ML models can reach a more satisfactory estimation for the responsive surface function in comparison to the RSM. The GDBT has exhibited to outperform the RSM with an accuracy improvement for 50% on unseen experimental data.

show abstract

Response surface methodology for advanced manufacturing technology optimization: theoretical fundamentals, practical guidelines, and survey literature review

Cited by 80 publications

References 130 publications

A Study of the Electrodeposition of Gold Process in Iodine Leaching Solution

A Study of the Electrodeposition of Gold Process in Iodine Leaching Solution

Multi-objective optimization of heating system for rapid thermal cycling molding mold with internal induction heating

Introducing Machine Learning Models to Response Surface Methodologies

Contact Info

Product

Resources

About