Elsevier

Applied Surface Science

Volume 253, Issue 6, 15 January 2007, Pages 3254-3262
Applied Surface Science

Using response surface methodology for optimizing deposited partially stabilized zirconia in plasma spraying

https://doi.org/10.1016/j.apsusc.2006.07.021Get rights and content

Abstract

This paper presents the application of the surface response methodology and fractional factorial experiment in developing a robust partially stabilized zirconia (PSZ) coating in plasma spraying processes. In this study, a two-step experimental design with a nonlinear regression model is proposed. After conducting a series of initial experiments in a controlled environment, significant factors for plasma spraying processes are selected to construct an appropriate response surface model for developing a robust performance for plasma spraying processes.

The experimental results show that using a quadric response surface model with the proposed two-step design strategy is a simple, effective and efficient approach. In the optimized model, several response plots were generated to examine parameter effects on the profile hardness. Optimization of the PSZ performances in a plasma spray process has been performed.

Introduction

Thermal barrier coating is a relatively new technique that has come into use in the last two decades. It is used in several activities aimed at gas turbine engines with specific coating of superalloy materials to improve their efficiency. It also provides thermal resistance to the metallic substrate and reduces the substrate surface temperatures in many power generation systems, and hence improves the durability of the components in these systems. Instead of developing new bulk thermal resistance materials, thermal barrier coating is now the current trend for solving the engines’ thermal problems [1], [2]. It is ideal to find a specified superalloy coating on substrate matrix not only with low costs, but also with a good performance for high precision materials that are extremely difficult using conventional treatment processes. The plasma spray coating is a popular and efficient method of forming thermal barrier coatings. Compared to other ceramics, zirconia-based materials are preferred as coating materials in engineering applications due to its superior mechanical properties, such as anti-heat, low thermal conductivity, large coefficient of thermal expansion and stability, as well as good wear resistance.

Among the various coating materials of partially stabilized zirconia (PSZ), several binary and ternary oxide systems have also been developed as potential sources of hardening zirconia polycrystals. The application of zirconia-base with various stabilizing oxides such as MgO, CaO, CeO and Y2O3 is quite well known in thermal barrier coating that is made mostly by plasma spray. However, partially stabilized zirconia coating provides the most desirable properties in the anti-heat of these coatings and, therefore, is widely used. Zirconia-based coating after thermal barrier treatments has been the subject of numerous investigations. Studies focused on the identification and quantification of phase transformations in plasma spray coatings have been done [3], [4], [5]. These characteristics are helpful in predicting the coating behavior under controllable plasma spray processes, but are not sufficient in finding the means of a systematically optimal coating. In thermal barrier coating processes, appropriate modeling of a process model is hardly efficiently used, nor is a systematic analysis of the relationship between independent variables and responses. Further, the impacts and importance of plasma spraying process factors on the surface coatings are still not well understood. Therefore, it is desirable to develop a systematic design in the plasma spraying process.

Due to the nature of the surface hardening process and the noise variation, it is typical and reasonable to present process characteristics with a nonlinear model. In this paper, we applied the response surface methodology (RSM) that helps in developing a suitable approximation for the true functional relationship between the independent variables and the response variable that may characterize the nature of the coatings [7], [8], [9], [12]. This paper illustrates the two-stage optimal procedures. In the first-stage, the analysis of variance (ANOVA) is efficiently utilized for selecting the most significant factors among numerous factors that are to be further analyzed in a response surface model. In the second-stage, a RSM model is obtained from the statistical fitting method. Including only the important factors, the RSM model could efficiently present an approximate functional relationship from the experimental data. A second-order RSM model is used to generate the response surfaces under various conditions. This study also demonstrates the procedure of applying statistical technologies to optimize plasma spray coating for effective productions.

Section snippets

Plasma spraying

Thermal spraying is the process of spraying molten metal onto a surface to form a coating. Pure or alloyed metal is melted in a flame or electric arc and atomized by a blast of compressed air. The resulting fine spray builds upon a prepared surface to form a solid metal coating. Plasma spraying is a thermal spraying process in which a nontransferred arc of the gun is used to create an arc plasma for melting and propelling the surfacing material to the substrate. It is this process of applying a

The experimental design

Several experimental studies on the effects of two or more factors have been conducted [6]. It can be shown that factorial designs are usually most efficient for this type of experiment. A complete factorial design requires experiments including all possible combinations of the levels of the factors, which can be very expensive and time-comsuming. Therefore, fractional factorial designs can be used as an efficient alternative. A fractional factorial design requires running only a fraction of

Deposited micrographs

For the illustration purpose, Fig. 2 shows the magnified surface coating morphologies of the PSZ reinforced substrate coatings spray for Tests 6, 15 and 16 in Table 3. It can be seen that some partially melted and unmelted small particles are formed in the partially stabilized zirconia coatings because feedstock materials are heated and propelled as individual particles or droplets onto a surface [13]. Fig. 3 shows three typical cross-sectional SEM photographs of the plasma spray coatings

Concluding remarks

It was experimentally shown that the use of fractional factorial experiments coupled with statistical fitting model strategy is a simple, effective and efficient way to approach for developing robust, highly efficient, high quality plasma spray coating process. Optimization of processing variables and responses in the plasma spray coating process has been achieved through appropriate construction of the process model in RSM. Based on the experimental results, the following conclusions have been

Cited by (69)

  • Crystal structure and morphology of CeO<inf>2</inf> doped stabilized zirconia ceramics under high-frequency microwave field sintering

    2022, Ceramics International
    Citation Excerpt :

    The chemical doping method is to dope alkaline earth metal oxides (such as MgO, CaO) or rare-earth metal oxides (such as Y2O3, CeO2) in ZrO2 [26–29]. The solubility of these doped metal cations in ZrO2 is very high, and these metal cations can replace the position of Zr4+ after high-temperature treatment [30,31]. To maintain the electric neutrality of the material locally, oxygen vacancies are introduced and distributed in ZrO2 [32].

  • Digital transformation of thermal and cold spray processes with emphasis on machine learning

    2022, Surface and Coatings Technology
    Citation Excerpt :

    When the relationship between the experimental response and factors is not linear the response surface methodology (RSM) is used [89,90]. Methods such as the analysis of variance (ANOVA) for determining the significance of the regression coefficients (experimental factors) are presented in [91–94]. Many other studies are also reported on the application of statistical techniques in surface coating methods [95–98].

  • Optimisation on the stability of CaO-doped partially stabilised zirconia by microwave heating

    2021, Ceramics International
    Citation Excerpt :

    By controlling the quantity and type of stabilisers, ZrO2 can be divided into three types of materials from the perspective of microstructure with different stable states doped and crystal size. The mentioned types of ZrO2 are tetragonal zirconia polycrystalline (TZP), fully stabilised zirconia (FSZ) and partially stabilised zirconia (PSZ) which is a vigorous developed toughening ceramic material contains the tetragonal phase of zirconia [12]. At present, the main preparation method of PSZ ceramic materials is the arc melting method which has high requirements on temperature and disadvantages including long operation time and large energy consumption.

View all citing articles on Scopus
View full text