| Summary: | 碩士 === 國立成功大學 === 航空太空工程學系 === 102 === Study of the optimal process of spraying array for the selective absorbing film of solar collector
Author:Hsieh Tsung-Min
Advisor:Chang Keh-Chin
Department of Aeronautics and Astronautics, National Cheng Kung University
SUMMARY
This study reports the optimization of a multiple-nozzle spray coating process to improve the uniformity of absorptivity (α) and emissivity (ε) distributions on thickness sensitive spectrally selective solar film using the Taguchi method. The control factors include (A) separation distance between nozzles, (B) height from the nozzles to the spraying surface, (C) conveyor speed, and (D) input pressures of the dual fluid spray nozzle (gas and liquid paint). The optimal combination of control factors is 13.5 cm separation for factor (A), 8 cm height for (B), 2.983 cm/s conveyor speed for (C), and gas/liquid pressure of 3.0/1.5 kgf/cm^2 for (D). This combination increased the S/N ratio of α from 24.29 (original factor combination) to 33.15 and the S/N ratio of ε from 19.32 to 25.23.
A field experiment by means of a solar test stand determines the correlations between the variance of temperature distribution on collectors and S/N ratios of α and ε as −0.9513 (highly correlated) and −0.6794 (moderately correlated), respectively.
Key words: Taguchi method, solar thermal, spectrally selective absorber, optimal process
INTRODUCTION
Located in subtropical regions, Taiwan has sufficient sunlight for solar energy applications throughout the year. The solar energy spectrum is mostly distributed in the wavelength range 250–3000 nm, and the emitted radiation of solar thermal systems in working temperatures (under 400 °C) is in the band of IR and FIR (wavelengths over 3000 nm). The concept of spectral selectivity is to improve the thermal conversion efficiency using a selective absorbing surface with high absorptivity in the band of solar radiation and low emissivity in the long-wave range[3].
One of the most used absorbing surfaces is made of thickness sensitive spectrally selective (TSSS) paint. The absorptivity and emissivity of the TSSS surface are very sensitive to the thickness of the film; both increase with increasing thickness[4]. Gunde predicted the optimum thickness for maximum spectral selectivity by theoretical computation: the highest spectral selectivity of carbon black TSSS paint occurs when thickness is approximately 3–4 μm[5]. In the results of a solar experiment by Yung-Chieh, a TSSS film with thickness of approximately 3±1 μm reached the highest temperature among films under same sun radiation conditions[6].
This study’s purpose is to enhance the uniformity of absorptivity (α) and emissivity (ε) distributions on TSSS film. The Taguchi method is an optimization approach commonly used in engineering to analyze the effects of different control factors on the level of a designated characteristic. Using this method, this study will classify the significance of control factors by analysis of variance (ANOVA) and find a combination of control factor levels that yields the best uniformity of quality characteristics (α,ε).
THE TAGUCHI METHOD
Problem Description
The spraying process in the domestic solar heat collector industry has a problem with quality control due to the instability of the handheld spraying process. The objective of optimization is to increase the uniformity of the carbon black solar selective absorber film made by an automatic multiple-nozzle spraying process.
Quality Characteristics
Quality characteristics are observed parameters that determine the quality of a process or product. In this study, absorptivity and emissivity are defined as the quality characteristics, both as a function of film thickness.
Control factors
There are many factors influencing the performance of the nozzle array spraying process. Four important factors were chosen as control factors: (A) separation distance between nozzles, (B) height from the nozzles to the spraying surface, (C) conveyor speed, and (D) input pressure of the dual fluid spray nozzle (gas/liquid paint). Each of these factors is set to three levels (Table 1).
Experiment
A full factorial experiment for the four control factors at three levels has total 3^4 factor combinations to determine the factor effects and interactions between factors. Omitting some information concerning the interactions between factors, the Taguchi method uses orthogonal arrays to test select combinations of factors. These comprise the minimum number of combinations sufficient to determine the factor effects. Accordingly, the L_9 (3^4 ) orthogonal table is selected (Table 2), which tests only nine combinations and determines all the effects of each individual factor.
Significance Test
By studying the experimental statistical data, we can calculate the factor effects on the quality characteristics and their S/N ratios and plot a response graph (Figure 1). The experimental errors can be calculated by variance decomposition. The significance of each factor can be confirmed by ANOVA (Table 3), and then control factors can be classified according to the results of the significance test.
SOLAR FIELD EXPERIMENTS
The purpose of solar field experiments is to clarify the relationship between the variation of the temperature distribution on solar film in an equilibrium state and the uniformity of the absorptivity and emissivity. We will calculate the coefficients of correlation between the RMS of temperature and S/N ratios of α and ε after a series of solar experiments.
Calibration of Thermocouples
All thermocouples were dipped into a thermostatic water bath with constant temperatures of 50, 60, 70, 80, and 90 °C, and a minimum of 500 samples were recorded for accurate calculation of the mean value and regulate the measurement value x to the reference value y (temperature of the water bath) by the least squares method. The correction function (1) and its coefficients are solved by (2) and (3).
(1)
(2)
(3)
After calibration, the bias error between thermocouples can be reduced.
RESULTS AND DISCUSSION
Process Optimization
After ANOVA and the significance test, factors A, B, and C are classified as significant factors to raise uniformity. Thus, the factor combination for optimal process can be chosen from the response graph (Figure 1):
A2 B1 C1 D1.
The original factor combination is
A3 B2 C2 D1.
The experimental results show that the S/N ratios of absorptivity and emissivity are raised from 24.29 dB (original) to 33.15 dB (optimal) and from 19.32 dB to 25.23 dB, respectively.
Correlation between Temperature RMS and S/N Ratios of α and ε
The results of solar experiments are plotted in Figure 2. As the uniformity of solar absorbing film increases, the variance of temperature distribution decreases. The coefficient of correlation between temperature RMS and S/N ratios of α and ε are -0.9513 and -0.6794, highly correlated with S/N ratio of α; moderately correlated with S/N ratio of ε.
CONCLUSIONS
The proportion of solvent in atomized droplets of TSSS paint determines uniformity in the mechanism of attachment when droplets collide with the substrate surface. A higher concentration of solvent can help droplets stretch wider and flatter from the effects of gravity. The input pressures of gas and liquid determine how well the droplets atomize, and the distance that droplets fly to the substrate is equal to the nozzle height. The longer a droplet flies, the more the solvent is volatilized in the flight to the substrate, which is why a lower nozzle height can lead to a higher uniformity.
The experiment to evaluate the heat collecting ability of a solar absorber is designed to allow absorption of solar radiation until the absorber reaches equilibrium, and the efficiency is defined by the final temperature. A better definition of efficiency for solar heat collection is to compare the heat transfer from the absorber to the working system with the solar radiation intensity at a specific working temperature. A preliminary design of the experiment is shown in Figure 3.
Table 1. Control factors and levels
Table 2. Orthogonal Table L_9 (3^4 )
Table 3. ANOVA Table (S/N ratio of emissivity)
Figure 1. Response graph for S/N ratio of emissivity
Figure 2. Results of solar experiments and corresponding S/N ratios
Figure 3. Preliminary design of the solar experiment
|
|---|
