CFD Analysis and Experimental Evaluation of the Effective Parameters on Paint Homogeneity in Mixing Tanks

Transcription

1 ISTP-16, 2005, PRAGUE 16 TH INTERNATIONAL SYMPOSIUM ON TRANSPORT PHENOMENA CFD Analysis and Experimental Evaluation of the Effective Parameters on Paint Homogeneity in Mixing Tanks N. K. Mohtaram*, M. Mostafaiyan **, E. D. Moghaddam** and F. Sharif** * Amirkabir University of Technology, Mahshahr, Iran ** Department of Polymer engineering, Amirkabir University of Technology, Tehran, Iran Corresponding author: Sharif@aut.ac.ir, (+9821) , (+9821) Keywords: Computational fluid dynamics (CFD); Mixing operation; Design parameters; Paint Homogeneity Abstract The aim of this study is to evaluate the effects of design and operation parameters on the quality of mixing for paint and solvent. As the first step, viscosity measurement of the mixture is employed to evaluate the mixing quality. Then the relationship between mixing operation and design parameters with the quality of mixing is studied experimentally and analytically. In CFD studies, a criterion for prediction of mixing quality is introduced and compared with experimental observations. Effect of different impeller angles, number of impellers, impeller location (off center), impeller speed and level of the fluid in the tank are then given. 1 Introduction Mixing process has been a subject of engineering interest for many years. The experimental studies and numerical simulations of mixing in chemical reactors, food industries and polymer processing are some examples of this interest [1, 2]. Through the years researchers have been trying to find more efficient methods to improve the quality of mixing process. It is necessary to understand the operation parameters and their effects on the quality of mixing or the homogeneity of the product to be able to design properly. Therefore, CFD simulations have been employed to model the flow in stirred tanks and understand the relationship between flow condition and mixing quality [3-9]. Many have studied the effects of geometry [4-6], operating condition [7] and rheological behavior of the fluid on quality of mixing [8, 9], and also to predict the mixing time and to prepare the homogenization curves [10]. In this paper, we will discuss the laminar flow mixing; however there are many studies on turbulent flow mixing. As an example, Akiti and Armenante studied chemical reaction in stirred tanks with turbulent flow [1]. In this work, mixing of the paint and solvent in a stirred tank is studied. Physical and rheological characteristics of the paint and solvent are given. We also present geometrical consideration. As this study is performed for Iran Khodro Industrial Group (a car producing company in Iran,I.K.I.G), the geometry of the mixing system is based on one of the stirred tanks that are used for paint homogenization. Viscosity measurements that were used for experimental evaluation of mixing are then reported. Results from CFD analysis and its comparison with experimental data are then reported. Finally, mixing operation parameters and mixer design parameters such as impeller angles, number of impellers, location of 1

2 N. K. Mohtaram, M. Mostafaiyan, E. D. Moghaddam, F. Sharif Fig.1. Stirred tank and impeller shape as used in simulation impeller (Off center), impeller speed and level of the fluid in the tank are studied. 2 Material Specifications In automobile industries, paint is sprayed on the cars requiring low viscosity. Thus it must be diluted by an appropriate solvent as practiced in paint preparation section of I.K.I.G in stirred tanks. The characteristics of the paint and its solvent which are considered in this study are given in Table 1. Component Paint Solvent Density kg/m3 kg/m 3 Viscosity mpa.sec mpa.sec Mass fraction Table 1. Paint properties and specification 3 Mixing Tank and dimensions In this part, a stirred tank which is used to prepare diluted paint is given. The Geometry of the stirred tank is presented in Fig.1. The impeller is off-center and the distance from main axis is 10 cm. Tank diameter is 93 cm and the height of the tank is 101 cm. Length of the flat blade pitched paddle and the width is 24.3 and 8.5 cm consequently. The over lab length is 12 cm and the angle is blade impeller length is 12.5cm and the width is roughly 2.4 cm. The angle of impeller in this type is Experimental Evaluations For our experiments, viscosity of the fluid is evaluated by ford cup. The liquid drop time ( T ) from the ford cup is directly related to the viscosity of the fluid ( η ). For higher viscosities the drop time increases. T ~ η (1) The drop time is measured by taking samples from the top of the stirred tank. In the mixing process, the paint is distributed in the solvent, and the viscosity of the fluid increases to an asymptotic value which indicates the mixing goodness and completion of the mixing process. Therefore G(t) is defined as a dimensionless number that indicates the progress of mixing goodness in terms of drop time. G () t T Tt = (2) T T 0 2

3 CFD ANALYSIS AND EXPERIMENTAL EVALUATION OF THE EFFECTIVE PARAMETERS ON PAINT HOMOGENEITY IN MIXING TANKS T is the final value of the drop time, T t is the measured drop time at any time t, and T 0 is the drop time at the beginning of mixing. Using Eq. (1) and Eq. (2), G (t) is given as: η η () ( t) G t = (3) η η is the viscosity of the mixture at the time t and η 0 is the pure solvent viscosity. As the mixing process continues, η ( t) reaches η and G (t) approaches to zero showing better mixing quality. η 0 η is the viscosity of the final mixture, ( t) In order to evaluate our work, we compared experimental and numerical results, using G (t), in Fig.2. 5 Numerical Analysis For the purpose of numerical analysis the two fluids (paint and solvent) were assumed completely separated at the beginning of the mixing process. No slip boundary conditions were applied at the tank wall and constant rotational speed was applied at the impeller. At the top of the tank atmosphere pressure is applied. Finite volume analysis was used for this study. The value of G (t) is also defined for our numerical method, employing Eq. (3). The Eq. (3) is computed from volumetric integral of viscosity over the whole domain of the stirred tank. 6 Results and Discussions Results from experimental and numerical analysis are compared in this section, to show the validity of the analysis in predicting mixing progress. Then the effects of different operating and design parameters including impeller angle, number of impellers, location and speed of impellers and level of fluid in the tank are discussed. Fig.2. Compared experimental and numerical results 6.2 Parametric Study To study the effects of different parametric on mixing process, a simple rectangular impeller has been chosen. The dimension of the impeller is shown in Table 2. Impeller Length 10 cm Impeller width 5 cm Impeller thickness 0.5 cm Table 2. Impeller Dimensions Also we have chosen a base set of operation parameters. These parameters are described below. Impeller velocity 15 rad/sec Mixing time 1 hr Impeller angle 15 degrees Impeller quantity 1 Table 3. Base set of operation parameters 6.1 Comparison of experimental and numerical results 3

4 N. K. Mohtaram, M. Mostafaiyan, E. D. Moghaddam, F. Sharif Impeller angles In this section the performance of a single impeller with different angles relative to agitation main axis is considered. The simulation results are illustrated by Fig.3. Fig.4 Simulation results of mixing performance for Single mixer (A) and, double mixer (B) Impeller location Fig.3. Simulation results for mixing performance of impellers with various angles Fig.5 presents the results for the effect of location of an impeller, being at the center or offcenter in the tank. Fig.3 shows that the impeller with the angle of 30 degrees results in faster homogenization Number of impellers Mixing performance is also influenced by the number of impellers. Fig.4 illustrates the difference between the performance of presents single (A) and double (B) impellers. As shown in Fig.4 for the case of single impeller, the mixing is not complete even after one hour. For the case of double impeller, however; mixing is nearly complete after forty minutes. Fig5.Simulation results of mixing performance due to impeller location 4

5 CFD ANALYSIS AND EXPERIMENTAL EVALUATION OF THE EFFECTIVE PARAMETERS ON PAINT HOMOGENEITY IN MIXING TANKS The off-center location results faster in better quality. It must be noted that off center location requires ore energy but this study provides quantitative measures for choosing better design Impeller speed The effect of impeller speed is illustrated in Fig.6. The figure shows that increasing the speed of impeller can greatly improve quality of mixing. Fig.7 Simulation results of mixing performance due to level of fluid in the tank Fig.6 Simulation results of mixing performance due to impeller speed Level of the fluid in the tank The influence of the level of the fluid in stirred tank in order to evaluation mixing performance is presented in Fig.7. 7 Conclusions In this work mixing of paint in a solvent was studied. It was illustrated that there are many geometrical parameters, and operating parameters which are involved in determining the quality of the mixture and the mixing time. Impeller angle, quantity of impeller, impeller location, impeller speed, and the level of the fluid in the tank are the parameters considered in this study. Effect of these parameters were studied and quantified by comparing them to a reference set of parameters, and it was shown that changing the geometrical and the operating parameters can improve the mixing process. Acknowledgement The support of Iran Khodro Industrial Group (IKIG) regarding the experimental measurements is gratefully acknowledged. The authors would also like to thank Paint Engineering Staff of IKIG for their sincere help. 5

6 N. K. Mohtaram, M. Mostafaiyan, E. D. Moghaddam, F. Sharif References [1]. Akiti O and Armenante P, A computational and experimental study of mixing and chemical reaction in a stirred tank reactor equipped with a down-pumping hydrofoil impeller using a micro-mixing based CFD model, 10 th European Conference on Mixing, Delft, Netherlands, [2]. Middleman S, Fundamental of polymer processing. McGraw-Hill, Inc [3]. Lane g, Schwarz M and Evans M. Comparison of CFD methods for modeling of stirred tanks. 10 th European [4]. Montante G, Micale G, Brucato A and Magelli F. CFD simulation of particle distribution in a multipleimpeller high aspect ratio stirred vessel. 10 th European [5]. Syrjänen J, Manninen M. Detailed CFD prediction of flow around a 45 pitched blade turbine. 10 th European [6]. Musgrove M and Ruszkowski S. Influence of impeller type and agitation conditions on the drop size of immiscible liquid dispersions. 10 th European [7]. Taşkin G and Wei H. The effect of impeller-to-tank diameter ratio on draw down of solids, Chemical Engineering Science, 58 (2003) [8]. Ulbrecht J, Mixing of viscous non-newtonian liquids. State University of New York,1998. [9]. White S, Polymer Mixing. 1st edition, Hanser, [10]. Giuseppina M, Moštěk M, Jahoda M and Mangell, F. CFD simulations and experimental validation of homogenization curves and mixing time in stirred Newtonian and psudoplastic liquids, Chemical Engineering Science, Article In Press, Accepted 1 Nov