Advances in Sustainable Petroleum Engineering

Advances in Sustainable Petroleum Engineering Science Volume 5, Number 2, 2013 Table of Contents A Modified Version of the Aziz et al. Multiphase Flow Correlation Improves Pressure Drop Calculations in High-Rate Oil Wells Hazim H. Al-Attar, Mohamed Y. Mohamed, and Mohamed E. Amin The Synergetics of Aggregation Processes and Interfacial Phenomena in Polypropylene/Calcium Carbonate Nanocomposites G. M. Magomedov, Kh. Sh. Yakh’yaeva, G. V. Kozlov, and G. E. Zaikov Experimental Study of Wax Deposition in Single-Phase Subcooled Oil Pipelines O. A. Adeyanju and L. O. Oyekunle Effect of the Modification by Organic Silicon Substances of the Mineral Fillers on Some Properties of Epoxy Resin Composites J. Aneli, O. Mukbaniani, E. Markarashvili, G. Zaikov, and E. Klodzinska

55

69

79

97

Surface Activation of Fibrous PET Materials N. P. Prorokova, A. V. Chorev, S. M. Kuzmin, S. Yu. Vavilova, and V. N. Prorokov

105

Modeling the Movement of Dust Particles in the Swirling Flow R. R. Usmanova and G. E. Zaikov

121

New York

Advances in Sustainable Petroleum Engineering Science Advances in Sustainable Petroleum Engineering Science aims to demystify the science behind energy sustainability, with particular focus to petroleum engineering. Unlike other journals, ASPES focuses on fundamental science, long-term research, and true sustainability. This Journal offers a true paradigm shift, which has the potential of reverting the current yoyo culture of implosive technology development. Because sustainability requires the enforcement of natural laws at all aspects of development, all topics of petroleum engineering will be entertained. Editors-in-Chief Khalid Aziz, Stanford University, U.S.A. Turgay Ertekin, Pennsylvania State, University, U.S.A. Rafiq Islam, Dalhousie University, Canada Managing Editor Chefi Ketata, Consultant Contact email: [email protected] Editorial Board Members Gennady E. Zaikov, Russian Academy of Sciences Advances in Sustainable Petroleum Engineering Science is published quarterly by Nova Science Publishers, Inc. 400 Oser Avenue, Suite 1600 Hauppauge, New York 11788-3619, U.S.A. Telephone: (631) 231-7269 Fax: (631) 231-8175 Email: [email protected] Web: www.novapublishers.com ISSN: 1937-7991 Print: $370

Subscription Price per Volume (2013) Electronic: $370 Combined Print + Electronic: $555

Additional color graphics might be available in the e-version of this journal. Copyright © 2013 by Nova Science Publishers, Inc. All rights reserved. Printed in the United States of America. No part of this Journal may be reproduced, stored in a retrieval system, or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical, photocopying, recording, or otherwise without permission from the Publisher. The Publisher assumes no responsibility for any statements of fact or opinion expressed in the published papers.

Advances in Sustainable Petroleum Engineering Science Volume 5, Number 2

ISSN: 1937-7991 © Nova Science Publishers, Inc.

A MODIFIED VERSION OF THE AZIZ ET AL. MULTIPHASE FLOW CORRELATION IMPROVES PRESSURE DROP CALCULATIONS IN HIGH-RATE OIL WELLS Hazim H. Al-Attar, Mohamed Y. Mohamed, and Mohamed E. Amin United Arab Emirates University, Al Ain, United Arab Emirates

ABSTRACT The prediction of multiphase pressure drop during the simultaneous flow of oil, gas and water in vertical tubing strings is crucial in the development and optimum exploitation of an oil field. In this paper a method is proposed to improve pressure drop predictions by Aziz et al. multiphase vertical-flow correlation. This correlation is theoretically justified as compared to the traditional empirical methods. The present method suggests combining several flow pattern maps with the Aziz et al. correlation in an attempt to achieve the improvement sought. Two field data sets comprising 32 production tests gathered from three different sources in the Middle East and North Africa were used to examine the performance of the various combinations. The results of this work indicate that the performance of Aziz et al. multiphase correlation can be best improved by replacing its original flow-pattern map with the traditional Duns-Ros flow-regime map and for both data sets used. A significant improvement has been observed giving an overall absolute average percent deviation of 2.16% compared with 5.33% for the original correlation. Also, with this combination the relative performance factor has been reduced to 1.33 from 2.90 for the original Aziz et al. correlation. The efficiency of the Aziz et al.-Duns and Ros combination at high oil production rates, when compared with the original Aziz et al. correlation, was further examined by noticing the improvement gained in statistical measures at higher levels of liquid superficial velocity. This work represents an addition to the technology of multiphase behavior in vertical pipes and its results will help in a more accurate design of tubing strings in high-rate producing oil wells.

Keywords: Multiphase Flow, Vertical Wells, Pressure Drop, High-Rate, Aziz Correlation 

E-mail: [email protected]

56

Hazim H. Al-Attar, Mohamed Y. Mohamed, and Mohamed E. Amin

NOMENCLATURE d = tubing ID, ft eri,ei = statistical measures defined in eqns. (2) and (6), respectively statistical parameters defined in eqns. (1), (3), (4), (5), (7), and (8), respectively g = acceleration of gravity = 32.2 ft/sec2 LB = 1.071 – (0.2281 vm2/d) Lm = 75 + 84 NLv 0.75 Ls = 50 + 36 NLv L1, L2 = function of Nd n = number of data points in the statistical sample Nd = pipe-diameter number, dimensionless Ngv = gas-velocity number, dimensionless NLv = liquid-velocity number, dimensionless vm = average mixture velocity, fps vs = slip or bubble-rise velocity, fps vsg = gas superficial velocity, fps vsL = liquid superficial velocity, fps pc = calculated pressure drop, psi pm = measured pressure drop, psi g = density of gas phase, lbm/ft3 L = density of liquid phase, lbm/ft3 L = surface tension of liquid phase, dynes/cm

SI Metric Conversion Factors °API 141.5/(131.5 + °API ) = g/cm3 Bbl x 1.589873 E–01 = m3 ft x 3.048* E–01 = m °F (°F – 32)/1.8 = °C in. x 2.54* E+00 = cm psi x 6.894757 E+00 = kPa scf/stb x 1.801175 E–01 = sm3/stm3 *Conversion factor is exact

INTRODUCTION Predicting multiphase flow performance in vertical wells is crucial in the design of any production system (Brown and Lea, 1985). One of the most important components in the well system is the well tubing. As much as 80 percent of the total pressure loss can be consumed in lifting the reservoir fluids from the bottom of the hole to the surface (Beggs, 2008). Several methods have been proposed to predict gas-liquid mixture flow in wells. Because of the highly complex and unpredictable nature of multiphase flow, most early investigators used laboratory and/or field data to develop empirical correlations for evaluating pressure

A Modified Version of the Aziz et al. Multiphase Flow Correlation …

57

drop during multiphase flow (Poettmann and Carpenter, 1952, Baxendell and Thomas, 1961, Fancher and Brown, 1963, Hagedorn and Brown, 1965, Asheim, 1986, Duns and Ros, 1963, Orkiszewski, 1967, Aziz et al., 1972, Chierici et al., 1974, Beggs and Brill, 1973 and Mukherjee and Brill, 1985). The validity of these empirical correlations is to some degree limited to the quality and scope of the data and the type of experimental measurements used in their development. Lawson and Brill (1973) examined the accuracies of these empirical correlations and concluded that no single pressure-loss correlation was found superior to all others considered for all ranges of producing well flow parameters. Pucknell et al. (1993) compared predicted pressure drops with observed pressure drops for 246 data sets collected from eight fields and concluded that despite the development of new mechanistic models, no single method gives accurate predictions of bottomhole flowing pressures in all fields. They also concluded that traditional methods such as Duns and Ros method (1963) give good results in oil wells and that the Ansari et al. (1994) mechanistic model gives the best results of all methods evaluated. Salim and Stanislav (1994) compared methods that describe the flow of gas-liquid mixtures in vertical wells with 189 data sets gathered from five different sources, representing annular-mist flow pattern, and concluded that the empirical correlations by Orkiszewski (1967) and Duns-Ros appeared to be less accurate than mechanistic models. Based on 94 field tests collected from a single source in vertical wells, Al-Attar et al. (2011) concluded that for vertical multiphase flow the more reliable methods were the combination of Mukherjee-Brill (1985) and Beggs-Brill (1973) correlations, followed by the Ansari et al. mechanistic model (1994), Duns and Ros (1963) correlation, and Hagedorn and Brown (1965) correlation. They also concluded that bottom to top calculations give more accurate predictions of pressure drops than top to bottom calculations and that the former scheme of calculations should be used whenever possible. Late investigators have recognized that improved understanding of multiphase flow in pipes required a sophisticated combined experimental and theoretical approach (Brill and Mukherjee, 1999). This understanding was transformed into improved mechanistic models to better describe the physical phenomena occurring (Ozon et al., 1987, Hasan and Kabir, 1988, Ansari et al., 1994, Xiao et al., 1990, Chokshi, 1994 and Chokshi et al., 1996). Among the empirical correlations, the Aziz et al. (1972) seems to have some theoretical justification and with some modification it could predict multiphase flow performance inside production tubings more accurately. In this paper a method is proposed for improving the ability of Aziz et al. correlation to predict pressure drops in high-rate oil producing vertical wells. It is suggested here to replace the original Aziz et al. flow pattern map by other selected flow pattern maps which were developed by Duns and Ros (1963), Orkiszewski (1967), Ansari et al. (1994), and Hasan and Kabir (1988), respectively.To achieve this objective, 32 field test measurements representing two sets of published data gathered from the Middle East (Al-Attar et al., 2011) and North Africa (Al-Attar and Abu-Ghalia, 1995) were used in the calculations of pressure drops. All calculations were performed with a popular industry-standard software package.

Field Observations A total of 32 published field tests were considered in this study. These data represent two sets of observations taken from two different geographical regions. The first set comprises

58


twenty nine tests taken from six wells that have been producing from a prolific oil field in the Middle East (Al-Attar et al., 2011). The second set of data consists of three tests taken from three wells which have been producing from two major oil fields in North Africa region (AlAttar and Abughalia, 1995). Table 1 illustrates the ranges of flow parameters of these data. It should be mentioned that the six wells completed in the Middle East field are partially tubed (non-uniform completion) and flowed on annulus. Also, two flow regimes have been identified in these six wells, i.e., single-phase in the lower part of the well and two-phase flow in the upper part of the well. Table 1. Ranges of flow parameters of Data Sets Flow Parameter Oil Flow Rate, stb/d Gas Oil Ratio, scf/stb Water Oil Ratio, stbw/stbo Water specific gravity (fresh water = 1.0) Tubing ID, inch Casing ID, inch Depth, ft Tubing Head Flowing Pressure, psi Bottomhole Flowing Pressure, psi Tubing Head Flowing Temperature, °F Bottomhole Flowing Temperature, °F Tank Gravity of Oil, degrees API Tank Gas Specific Gravity (air = 1.0)

From 416 475 0 1.131 2.441 4.778 5900 440 1660 100 200 32 0.738

To 29425 2100 0.11 1.157 3.500 6.094 15130 1817 5540 220 320 40 1.055

METHODOLOGY Selection of the candidate correlation for further improvement was based on two criteria: (1) the original development of the various correlations and (2) statistical results on deviations between measured and predicted pressure drops by the various correlations. For this purpose the most popular correlations used by the industry and are included in almost every commercial software package were considered to predict pressure drops inside each well. The correlations considered in this study are the ones developed by Poettmann and Carpenter (1952), Hagedorn and Brown (1965), Duns and Ros (1963), Orkiszewski (1967), Aziz et al. (1972) and Ansari et al. mechanistic model (1994), respectively. Reliable PVT correlations (McCain, 1991) were consistently implemented in all pressure drop calculations by the various correlations. The calculations of bottom-hole flowing pressures or tubing-head flowing pressures by the various methods were performed by a popular industry-standard software package. This software offers a variety of options regarding vertical lift correlations, PVT empirical correlations, well completion configurations, temperature profiles, bottom to top and top to bottom calculations and effects of flow variables (sensitivity analysis). After the candidate multiphase flow correlation has been selected, the next step was to improve its performance by combining it with several published flow pattern maps. In this


59

work four of the most widely used maps have been attempted, two of which represent the empirical approach and two represent the mechanistic modeling approach for predicting pressure drops in vertical tubing strings. The selected flow pattern maps of the former approach are: (1) Duns and Ros (1963) and (2) Orkiszewski (1967), and of the latter approach are: (1) Ansari et al. (1994) and (2) Hasan and Kabir (1988).The boundary locations proposed by these authors in their corresponding flow pattern maps are given in the Appendix. The tubing pressure drops predicted by the original candidate correlation and by its modified versions were then compared with the measured values and statistical deviations reported.

RESULTS AND DISCUSSION The statistical parameters considered in this study are defined as follows (Brill and Mukherjee, 1999). E1 = [(1/n) ∑eri] x 100

(1)

where eri = [(pc - pm) / pm]

(2)

E1 indicates the overall trend of the performance relative to the measured pressure drop (average percent difference). E2 = [(1/n) ∑│eri│] x 100

(3)

E2 indicates an average of how large the errors are (absolute average percent difference). E3 = ∑ [(eri – E1) 2 / (n – 1)] 0.5

(4)

E3 indicates the degree to which the errors are scattered about their average percent error (standard deviation). E4 = [(1/n) ∑ei]

(5)

where ei = pc - pm

(6)

E4 indicates the overall trend independent of the measured pressure drop. E5 = [(1/n) ∑│ei│]

(7)

E5 also is independent of the measured pressure drop and indicates the magnitude of the average error.

60

Hazim H. Al-Attar, Mohamed Y. Mohamed, and Mohamed E. Amin E6 = ∑ [(ei – E4) 2 / (n – 1)] 0.5

(8)

E6 indicates the scattering of the errors about their average error. The subscript i in the above equations vary between 1 and n, where n is the number of data points in the statistical sample. The evaluation of the various models used in this study was accomplished by a comparison of the above statistical parameters. The evaluation also involved the use of a relative performance factor, which is defined by Brill and Mukherjee, 1999: Frp = [(│E1│−│E1min│) / (│E1max│−│E1min│)] + [(E2 – E2min) / (E2max – E2min)] + [(E3 – E3min) / (E3max – E3min)] + [(│E4│−│E4min│) / (│E4max│−│E4min│)] + [(E5 – E5min) / (E5max – E5min)] + [(E6 – E6min) / (E6max – E6min)]

(9)

It should be mentioned at this point in this investigation that pressure drops predicted from bottom to top have been found more accurate than pressure drops calculated from top to bottom which confirms Gregory et al., 1980 observations and Al-Attar et al. (2011) findings. The statistical results for the various prediction methods when applied to all 32 well tests from bottom to top are listed in Table 2. The minimum and maximum possible values for the relative performance factor, Frp, are 2.90 and 5.37, indicating the best and worst performances, respectively. These results indicate that the Aziz et al. correlation seems to predict pressure drops better than the other correlations selected in this study. Table 2. Statistical evaluation of the selected correlations Correlation Aziz et al. Duns and Ros Orkiszewski Ansari et al. Hagedorn and Brown Poettmann and Carpenter

E1 0.56 1.25 -0.73 1.04 -1.75 -4.91

E2 5.33 4.14 5.19 4.27 5.00 6.68

E3 3.20 4.07 4.13 5.90 9.95 27.93

Frp 2.90 2.93 4.26 3.18 3.52 5.37

In addition to the various statistical results presented in Table 2, further information about the performance of the selected pressure drop prediction methods is shown in Figure 1 and Figure 2. These figures illustrate the effect of liquid production rate (superficial liquid velocity) on the performance of each prediction method. This specific flow parameter is chosen not only because it is important in multiphase flow, but also because it represents the target of our investigation. It is expected that an analysis of prediction errors based on ranges of this variable should point out some of the strengths and weaknesses of the individual selected correlations. The superficial liquid velocity (vsL) is defined as the liquid production rate in stb/d divided by pipe cross-sectional area multiplied by the appropriate constant to give units of ft/sec (fps). This flow parameter is not normally used by petroleum production engineers, but was considered in this work to highlight trends based on the liquid production rate/tubing area ratio.


61

As illustrated in Figure 1 and Figure 2, the pressure drop predictions generally improve as superficial velocity of produced liquids increases. It seems that as production rates increase, the gas and liquid have tendency to flow more uniformly (no gas slippage conditions) than at lower liquid rates (with gas slippage). The Ansari et al. mechanistic model (which adapts flow pattern map developed by Taitel et al., 1980), Aziz et al., and Duns-Ros correlations seem to share the best performance among the selected prediction methods. For the above reasons it was decided to modify the Aziz et al. correlation to improve its predictive efficiency of pressure drops in high-rate producing oil wells. The modification of Aziz et al. correlation was accomplished by combining this correlation with different flow pattern maps. A summary of the statistical evaluation results for the candidate correlation before and after modification for all well tests is given in Table 3. The minimum and maximum possible values for the relative performance factor, Frp, are 1.33 and 2.90.

Figure 1. Standard deviation results for well data grouped by superficial velocity of produced liquids; all methods.

The results listed in Table 3 indicate that all modifications attempted on the original correlation did improve its performance but to varying extents. These improvements could be the result of the better identification of the boundaries separating the various flow regimes provided by the flow pattern maps attempted in this work. The best improvement was found with the combination of Aziz et al. correlation and Duns-Ros flow pattern map. A significant improvement has been observed with this combination giving an overall absolute average percent deviation of 2.16% compared with 5.33% for the original correlation. Also, with this combination the relative performance factor has been reduced to 1.33 from 2.90 for the original Aziz et al. correlation. A comparison between the calculated and measured pressure drops for the original Aziz et al. method and the combined Aziz et al. – Duns-Ros correlation

62


is illustrated in Figure 3. In this figure the two correlations are comparable at low pressure drops (low production rates) but at high pressure drops (high production rates) the modified correlation performance is clearly better than that of the original Aziz et al. correlation.

Figure 2. Average percent difference results for well data grouped by superficial velocity of produced liquids; all methods.

Table 3. Summary of statistical results before and after modification Method Original Aziz et al. Aziz et al. combined with Duns-Ros flow-pattern map. Aziz et al. combined with the criteria developed by Orkiszewski Aziz et al. combined with Ansari et al. flow-pattern map. Aziz et al. combined with the boundary locations proposed by Hasan and Kabir.

E1 0.56 0.45 0.04 0.27

E2 5.33 2.16 4.36 4.14

E3 3.20 3.13 1.32 1.81

Frp 2.90 1.33 2.76 1.95

0.32

3.27

1.83

2.08

The lower improvement achieved by the original Aziz et al.- Ansari et al. combination shown in Table 3 may be attributed to the fact that the Ansari et al. mechainistic model works significantly better than all other methods in the annular flow regime (Brill and Mukherjee, 1999). Analysis of tests data used in this study revealed that they predominantly fall in the domain of bubble flow regime and slug flow regime. However, as indicated in Table 3, the overall performances of all selected methods are comparable and more field data is needed to support the findings of this work.


63

Figure 3. Predicted vs. measured pressure drops as depicted by the original and modified Aziz et al. correlations.

In an attempt to further examine the superiority of the combined Aziz. et al. original correlation and the Duns-Ros map at high oil production rates, Table 4 was prepared to show the results of calculations of statistical measures in terms of vsL-groups for the various combinations attempted in this work. The best values of these statistical measures are being boldfaced indicating the improvement of the results at higher oil production rate and Figure 4 illustrates this trend.

Figure 4. Average percent difference for well data grouped by superficial velocity of produced liquids; original Aziz et al. vs. Aziz et al. combined with Duns and Ros map.

Table 4. Statistical results of the original Aziz et al. method and its modified versions based on vsL groups

E1

E3

Aziz et al. Combined with Duns-Ros Flow Pattern Map E1 E3

0.1-3.0 (5)

12.4

10.25

14.7

12.10

12.74

10.55

13.32

10.82

13.01

10.52

5.0-7.0 (6)

-1.70

3.25

-1.33

3.22

-0.76

3.75

-0.50

3.74

-0.35

4.05

7.0-8.0 (10)

-3.59

1.70

-3.37

1.68

-3.87

1.69

-3.59

1.71

-3.58

1.82

8.0-10.0 (5)

-1.28

3.47

-1.13

3.31

-2.77

4.95

-2.50

4.95

-2.59

4.94

> 10.0 (6)

1.41

3.44

-0.36

5.00

-1.12

5.13

-0.82

5.08

-0.92

5.11

vsL Group, fps (no. of data points)

Original Aziz et al. Correlation

Aziz et al. Combined with Orkiszewski Criteria E1 E3

Aziz et al. Combined with Hasan-Kabir Flow Pattern Map E1 E3

Aziz et al. Combined with Ansari et al. Flow Pattern Map E1 E3


65

CONCLUSION Based on the results of this study and the data sets used to produce these results, the following conclusions may be drawn: 1. The performance of the original Aziz et al. correlarion in high-rate producing oil wells can be improved when combined with other flow pattern maps. Four of the most widely used maps have been attempted, two of which represent the empirical approach and two represent the mechanistic modeling approach for predicting pressure drops in vertical tubing strings. These four maps are found to compare favorably in terms of the improvement sought. 2. The best performance of the attempted combinations is observed with the Duns and Ros flow pattern map. This observation was confirmed by various statistical measures including the relative performance factor. 3. The superiority of the Duns-Ros flow pattern map at high oil production rates was further examined and confirmed by running statistical evaluation in terms of the liquid superficial velocity. 4. It is expected that a comparison study like the present one would significantly rely on the quality and quantity of the data sets used. Consequently, further testing of the modified version of Aziz et al. correlation with data covering the various flow regimes encountered during multiphase flow in vertical wells is recommended.

APPENDIX Boundary locations proposed by: 1. Orkiszewski Criteria (1967) Bubble-Flow: vsg/vm < LB. Slug-Flow: vsg/vm > LB, Ngv < Ls. Mist-Flow: Ngv > Lm. Transition between Slug-Flow and Mist-Flow: Ls ≤ Ngv ≤ Lm. 2. Hasan and Kabir Mechanistic Model (1988) Bubble/Slug Transition: vsg = [sin /(4 – Co)](CovsL + vs) Slug/Churn Transition: vm1.12 = 14.265 d0.48 [g(L – g)/L]0.5 (L/L)0.6 (L/L)0.08 Annular-Flow Transition: vsg = 9.45[gL(L – g)/g2]1/4 3. Dun and Ros Criteria (1963) Region 1: 0 ≤ Ngv ≤ (L1 + L2NLv) Region 2: (L1 + L2NLv) ≤ Ngv ≤ Ls Region 3: Ngv > Lm Transition between Region 2 and Region 3: Ls < Ngv < Lm 4. Ansari et al. Mechanistic Model (1994) Bubble/Slug Transition: dmin = 57.94 {[(L – g)/L]/[L2g]}1/2 vsg = 0.762 vs + 1.105 vSL vs = 4.66 [gL(L – g)/L2]1/4 Dispersed Bubble/Bubble Transition:

66

Hazim H. Al-Attar, Mohamed Y. Mohamed, and Mohamed E. Amin 2{[0.4L/ g(L – g)] (L/L)3/5 (f/2d)2/5(vsL + vsg)1.2} = 0.725 + 4.15[vsg/ (vsg + vsL)]0.5 Transition to Annular Flow: vsg = 9.45[gL (L – g)/g2]1/4

REFERENCES Al-Attar, H., Abu Al-Soof, N. and Al-Ani, T.: “A Study of Multiphase Flow in High-Rate Oil Wells,” Journal of Advances in Sustainable Petroleum Engineering Science,” Volume 2, Issue 2 (2011), Pages 175-199. Al-Attar, H. and Abughalia, A.: “A Study of Vertical Multiphase Flow in Some Libyan Oil Wells”, Petroleum Research Journal,” Tripoli, Libya, Volume 7, 1995, Pages 1-30. Ansari, A. et al.: “A Comprehensive Mechanistic Model for Two-Phase Flow in Wellbores,” SPEPE (May 1994) 143; Trans., AIME, 297. Asheim, H.: “MONA, An accurate Two-Phase Well Flow Model Based on Phase Slippage,” SPEPE (May 1986) 221. Aziz, K., Govier, G., and Fogarasi, M.: “Pressure Drop in Wells Producing Oil and Gas,” J. Cdn. Pet. Tech. (July-Sept. 1972) 11, 38. Baxendell, P., and Thomas, R.: “The Calculation of Pressure Gradient in High-Rate Flowing Wells,” JPT (Oct. 1961) 1023; Trans., AIME, 222. Beggs, H. and Brill, J.: “A Study of Two-Phase Flow in Inclined Pipes,” JPT (May 1973) 607; Trans., AIME, 255. Brill, J. and Mukherjee, H.: Multiphase Flow in Wells, Monograph Volume 17, SPE, Richardson, Texas (1999). Chierici, G., Ciucci, G., and Sclocchi, G.: “Two-Phase Vertical Flow in Oil Wells – Prediction of Pressure Drop,” JPT (Aug. 1974) 927; Trans., AIME, 257. Chokshi, R.: “Prediction of Pressure Drop and Liquid Holdup in Vertical Two-Phase Flow Through Large Diameter Tubing,”PhD dissertation, U. of Tulsa, Tulsa, Oklahoma (1994). Chokshi, R., Schmidt, Z., and Doty, D.: “Experimental Study and the Development of a Mechanistic Model for Two-Phase Flow Through Vertical Tubing,” paper SPE 35676 presented at the 1996 Western Regional Meeting, Anchorage, 22-24 May. Design and Gas-Lift Installations,” Drill. and Prod. Prac. (1952) 257. Duns, H. Jr. and Ros, N.: “Vertical Flow of Gas and Liquid Mixtures in Wells,” Proc., Sixth World Pet. Cong., Tokyo (1963) 451. Fancher, G. Jr. and Brown, K.: “Prediction of Pressure Gradients for Multiphase Flow in Tubing,” SPEJ (March 1963) 59; Trans., AIME, 228. Gregory, M., Aziz, M., and Fogarasi, M.: “Analysis of Vertical Two-Phase Flow Calculations, Crude Oil, Gas Flow in Well Tubing,” J. Cdn. Pet. Tech. (Jan.-March 1980) 86-92. Hagedorn, A. and Brown, K.: “Experimental Study of Pressure Gradients Occuring During Continuous Two-Phase Flow in Small-Diameter Vertical Conduits,” JPT (April 1965) 475; Trans., AIME, 234. Hasan, A. and Kabir, C.: “A Study of Multiphase Flow Behavior in Vertical Wells,” SPEPE (May 1988) 263; Trans., AIME, 285.


67

Lawson, J. and Brill, J.: “A Statistical Evaluation of Methods Used to Predict Pressure Losses for Multiphase Flow in Vertical Oil,” paper SPE 4267 presented at the 48th Annual Fall Meeting of the Society of Petroleum Engineers of AIME, Las Vegas, Sept. 30 – Oct. 3, 1973. McCain, W. Jr.: “Reservoir-Fluid Property Correlations – State of the Art,” SPERE (May 1991) 266. Mukherjee, H. and Brill, J.: “Pressure Drop Correlations for Inclined Two-Phase Flow,” J. Energy Res. Tech. (Dec. 1985) 107, 549. Orkiszewski, J.: “Predicting Two-Phase Pressure Drops in Vertical Pipes,: JPT (June 1967) 829; Trans., AIME, 240. Ozon, P., Ferschneider, G., and Chwetzoff, A.: “A New Multiphase Flow Model Predicts Pressure and Temperature Profiles in Wells,” paper SPE 16535 presented at the 1987 SPE Offshore Europe Conference, Aberdeen, 8-11 September. Poettmann, F., and Carpenter, P.: “The Multiphase Flow of Gas, Oil and Water Through Vertical Flow Strings with Application to the Design and Gas-Lift Installtions,” Drill. and Prod. Prac. (1952) 257. Pucknell, J., Mason, J., and Vervest, E.: “An Evaluation of Recent Mechanistic Models of Multiphase Flow for Predicting Pressure Drops in Oil and Gas Wells,” paper SPE 26682 presented at the 1993 Offshore European Conference, Aberdeen, 7-10 September. Salim, P. and Stanislav, J.: “Evaluation of Methods Describing the Flow of Gas-Liquid Mixtures in Wells,” J. Cdn. Pet. Tech. (January/February 1994) 33, 58. Taitel, Y., Barnea, D., and Dukler, A.: “Modelling Flow Pattern Transitions for Steady Upward Gas-Liquid Flow in Vertical Tubes,” AIChE J. (May 1980) 25, No. 3, 345-354. Xiao, J., Shoham, O., and Brill, J.: “A Comprehensive Mehanistic Model for Two-Phase Flow in Pipelines,” paper SPE 20631 presented at the 1990 SPE Annual Technical Conference and Exhibition, New Orleans, 23-26 September.



THE SYNERGETICS OF AGGREGATION PROCESSES AND INTERFACIAL PHENOMENA IN POLYPROPYLENE/CALCIUM CARBONATE NANOCOMPOSITES G. M. Magomedov1, Kh. Sh. Yakh’yaeva1, G. V. Kozlov1, and G. E. Zaikov2 1

2

Dagestan State Pedagogical University, Makhachkala, Russian Federation N.M. Emanuel Institute of Biochemical Physics of Russian Academy of Sciences, Russian Federation

ABSTRACT Disperse nanoparticles aggregates formation in polymer nanocomposites has been studied. It has been shown that this process obeys to the synergetics laws. The mentioned aggregates are formed by structure reproduction replicative mechanism. The similar laws are also valid for interfacial regions formation. Therefore the synergetics laws allow nanocomposites structural characteristics prediction.

Keywords: nanocomposite, aggregation, interfacial phenomena, synergetics, bifurcation

INTRODUCTION One of the main trends of nanoworld objects properties study is their features registration for on the basis of synergetics principles [1, 2]. At present it is known [3], that nanoparticles structure is defined by a chemical interactions between atoms nature forming them. The fundamental properties of nanoparticles, forming in strongly nonequilibrium conditions is their ability to: 1. structures self-organization by adaptation to external influence; 2. an optimal structure self-choice in bifurcation points, corresponding to preceding structure stability threshold and new stable formation;




70

G. M. Magomedov, Kh. Sh. Yakh’yaeva, G. V. Kozlov et al. 3. a self-operating synthesis (self-assembly) of stable nanoparticles, which is ensured by information exchange about system structural state in the previous bifurcation point at a stable structure self-choice in the following beyond it bifurcation point [3].

These theoretical postulates were confirmed experimentally. In particular, it has been shown [3-5] that nanoparticles sizes are not arbitrary ones, but change discretely and obey to synergetics laws. This postulate is important from the practical point of view, since nanoparticles size is the information parameter, defining surface energy critical level [3]. Let us consider these general definitions in respect to polymer particulate-filled nanocomposites [1], for which there are certain distinctions with the considered above criterions. As it is well-known [1], nanofiller aggregation processes in either form are inherent in all types of polymer nanocomposites and influence essentially on their properties. In this case, although nanofiller initial particles have size (diameter) less than 100 nm, but these nanoparticles aggregates can exceed essentially the indicated above boundary value for nanoworld objects [6]. Secondly, nanofiller particles aggregates are formed at the expence of physical interactions, but not chemical ones. Therefore the present work purpose is the study of the synergetics laws applicability for nanofiller aggregation processes and interfacial phenomena description in particulate-filled polymer nanocomposites on the example of nanocomposite polypropylene/calcium carbonate [7].

EXPERIMENTAL Polypropylene (PP) of industrial production mark Kaplen 01 030 with weight – average molecular weight Mw of (2-3)105 and polydispersity index of 4.5 was used as matrix polymer and nanodimensional calcium carbonate (CaCO3) in the form of compound mark Nano-Cal R-1014 (China) with particles size of 80 nm and mass contents of 1-10 mass % was used as nanofiller. Nanocomposites PP/CaCO3 were prepared by components mixing in melt on twin-screw extruder Thermo Haake, model Reomex RTW 25/42, production of German Federal Republic. Mixing was performed at temperature 463-503 K and screw speed of 50 rpm during 5 min. Testing samples were obtained by casting under pressure method on casting machine Test Samples Molding Apparate RR/TS of firm Ray-Ran (Taiwan) at temperature 483 K and pressure 43 MPa. The rester-type electron microscopy (REM) method was used for nanocomposites PP/CaCO3 structure study. Study objects were prepared in liquid nitrogen with the purpose of microscopic sections obtaining. The scanning electron microscope with autoemissive cathode of high resolution JSM7500F of firm JEOL (Japan) was used for microscopic sections surface images obtaining. Images were obtained in the mode of low-energetic secondary electrons, since this mode ensures the highest resolution. Uniaxial tension mechanical tests have been performed on the samples in the shape of two-sided spade with sizes according to GOST 112 62-80. The tests have been conducted on universal testing apparatus Gotech Testing Machine CT-TCS 2000, production of German Federal Republic, at temperature 293 K and strain rate ~ 210-3 s-1.

The Synergetics of Aggregation Processes and Interfacial Phenomena …

71

RESULTS AND DISCUSSION A particulate nanofiller particles aggregate size (diameter) Dagr estimation can be performed according to the following formula [6]:

 25.1D1 / 3  1 / 3 D agr   k r     2 agr  Wn   2    ,

(1) where k(r) is an aggregation parameter,  is distance between nanofiller particles, Wn is nanofiller mass contents in mass %. In its turn, the value k(r) is determined within the frameworks of strength dispersion theory with the help of the relationship [1]:

n  m 

GbB k r  ,

(2)

where n and m are yield stress in compression testing of nanocomposite and matrix polymer, accordingly, G is shear modulus, bB is Burders vector. The included in the equation (2) parameters are determined as follows. The general relationship between normal stress  and shear stress  has the following look [8]:



 3.

(3)

Young’s modulus E and shear modulus G are connected between themselves by the simple relationship [9]:

G

E df

,

(4)

where df is nanocomposite structure fractal dimension, which is determined according to the equation [9]:

d f  d  11   

,

(5)

where d is the dimension of Euclidean space, in which fractal is considered (it is obvious, that in our case d=3),  is Poisson’s ratio, which is estimated by the mechanical testing results with the help of the relationship [10]:

Y 1  2  En 61    ,

(6)

72

G. M. Magomedov, Kh. Sh. Yakh’yaeva, G. V. Kozlov et al.

where Y and En are yield stress and elasticity modulus of nanocomposite, respectively. Burgers vector value bB for polymeric materials is determined according to the equation [11]:

 60.5   bB   C   

1/ 2

,

(7)

where C is characteristic ratio, connected with df by the equation [11]:

C 

2d f

d d  1d  d f 



4 3

.

(8)

The calculation according to the equations (1)-(8) showed CaCO3 nanoparticles aggregates mean diameter growth from 85 up to 190 nm within the range of Wn=1-7 mass % for the considered nanocomposites PP/CaCO3. These calculation can be confirmed experimentally by the electron microscopy methods. In Figure 1 the nanocomposites PP/CaCO3 with nanofiller contents Wn=1 and 4 mass % sections electron micrographs are adduced. As one can see, if at Wn=1 mass % nanofiller particles are not aggregated practically, that is, their diameter is close to CaCO3 initial nanoparticles diameter (~ 80 nm), then at Wn=4 mass % the initial nanoparticles aggregation is observed even visually and these particles aggregates sizes are varied within the limits of 80-360 nm. The adduced above estimations correspond to the results of calculation according to the equation (1) at the indicated CaCO3 contents: 85 and 142 nm, accordingly. Hence, the considered above technique gives reliable enough estimations of nanofiller particles aggregates diameter. It has been shown earlier on the example of different physical-chemical processes, that the self-similarity function has an iteration type function look, connecting structural bifurcation points by the relationship [12]:

Am 

n  1i/ m  n1 ,

(9)

where Am is the measure of aggregate structure adaptability to external influence, n and n+1 are preceding and subsequent critical values of operating parameter at the transition from preceding to subsequent bifurcation point, i is the structure stability measure, remaining constant at its re-organization up to symmetry violation, m is an exponent of feedback type; the value m=1 corresponds to linear feedback, at which transitions on other spatial levels are realized by multiplicative structure reproduction mechanism and at m2 (nonlinear feedback) – replicative (with structure improvement) one.


73

Figure 1. Electron micrographs of sections of nanocomposites PP/CaCO3 with nanofiller mass contents Wn=1 (a) and 4 (b) mass %.

Selecting as the operating parameter critical value nanoparticles aggregates diameter Dagr [3] at successive Wn change, the dependence of adaptability measure Am on Wn can be plotted, which is shown in Figure 2. As one can see, for the considered nanocomposites the condition is fulfilled

Am 

Dagr i Dagr i 1

 const  0.899 (10)

with precision of 2 %. This means, that aggregation processes in the considered nanocomposites obey to the synergetics laws, although their aggregates size exceeds the boundary value of 100 nm for nanoworld [3]. Let us note the important aspect of the dependence Am(Wn), adduced in Figure 2. The condition Am=const is kept irrespective of gradation, with which Wn changes – 0.5 or 1.0 mass %.

Figure 2. The dependence of adaptability measure Am on nanofiller mass contents Wn for nanocomposites PP/CaCO3.

74


As it is known [13], an interfacial layer in polymer nanocomposites can be considered as a result of two fractal objects (polymer matrix and nanofiller particles surface) interaction, for which there is the only linear scale l, defining these objects interpenetration distance [14]. Since the filler elasticity modulus is, as a rule, considerably higher than the corresponding parameter for polymer matrix, then the indicated interaction comes to filler indentation in polymer matrix and then l=lif, where lif is interfacial layer thickness [1, 13]. In this case it can be written [14]:

D  lif  a agr   2a 





2 d  d surf / d

,

(11)

where a is a lower linear scale of fractal behavior, which for polymeric materials is accepted equal to statistical segment length lst [11], dsurf is nanofiller particles (aggregates of particles) surface fractal dimension. lst is determined according to the equation [15]:

lst  l0C ,

(12)

where l0 is the main chain skeletal bond length, equal to 0.154 nm for PP [16]. The dimension dsurf is calculated in the following succession. First the nanofiller particles aggregate density n is estimated according to the formula [1]:

n  188Dagr 

1/ 3

, kg/m3,

(13)

where Dagr is given in mcm and then the indicated aggregate specific surface Su is determined [17]:

Su 

6 n Dagr

.

(14)

And at last, the value dsurf calculation can be fulfilled with the help of the equation [1]:

D  Su  410 agr   2 

d surf  d

,

(15)

where Su is given in m2/g, Dagr – in nm. The calculation according to the offered technique has shown lif increase from 1.78 up to 5.23 nm at Wn enhancement within the range of 1-7 mass %. The estimations according to the

The Synergetics of Aggregation Processes and Interfacial Phenomena … equation (9), where as n and n+1 the values

lif n

and

lif n 1

75

were accepted, showed that the

following condition was fulfilled:

Am 

lif n lif n1

 0.880 (16)

with the precision of 7 %. Hence, an interfacial layers formation in polymer nanocomposites, characterizing interfacial phenomena in these nanomaterials, obeys to the synergetics laws with the same adaptability measure, as nanofiller particles aggregation. Nevertheless, it should be noted, that this analogy is not complete for the considered nanocomposites within the range of Wn=1-7 mass % the value Dagr increases in 2.24 times, whereas the value lif does almost in three times. Let us consider further the exponent m in the equation (9) choice, characterizing feedback type in aggregation process. As it has been noted above, this exponent is equal to 2 in case of aggregates structure improvement, which can be characterized by their fractal dimension

d agr f . This dimension can be calculated with the help of the equation [18]: D  n  dens  agr   2a 

d agr f d

,

(17)

where dens is massive material density, which is equal to 2000 kg/m3 for CaCO3, a is a lower linear scale of fractal behavior, accepted equal to 10 nm [19]. In Figure 3 the dependence of

d agr f

on CaCO3 mass contents Wn for the considered

nanocomposites is adduced. As one can see, within the studied range of Wn the essential

d agr f

growth (from 2.34 up to 2.73 at general

d agr f

variation within the limit of 2.0 to 2.95

[9]) is observed, that can be classified as nanofiller particles aggregates structure improvement, as a minimum, by two reasons: their disaggregation level reduction and critical structural defect sizes decreasing [20]. Therefore, proceeding from the said above, it should be accepted that m=2, which according to the equation (9) gives i=0.808. Let us note, that this i value defines very stable nanostructures. So, for self-operating nano-solid solutions synthesis the values i=0.2550.465 at m=2 were obtained and in addition it has been shown that an optimal technological regime indicator is i=0.465 attainment at nonlinear feedback (m=2) realization [4]. Let us consider in conclusion the possibility of CaCO3 nanoparticles aggregation process prediction within the frameworks of synergetics treatment. In Figure 4 the comparison of CaCO3 aggregates diameter values, calculated according to the equations (1) Dagr and (9) syn Dagr

at Am=const=0.899. As one can see, this comparison demonstrates very good

76


conformity of nanofiller particles aggregates diameter, calculated by both indicated methods (the average discrepancy of Dagr and

syn Dagr

makes up 2 %).

Figure 3. The dependence of nanofiller particles aggregates structure fractal dimension

d agr f

on its

mass contents Wn for nanocomposites PP/CaCO3.

Figure 4. The comparison of nanofiller particles aggregates diameter, calculated according to the equations (1) Dagr and (9)

syn Dagr , for nanocomposites PP/CaCO3.


77

CONCLUSION Hence, the present paper results have demonstrated that disperse nanoparticles aggregates formation in polymer nanocomposites obeys to synergetics laws even at these aggregates size larger than upper dimensional boundary for nanoparticles, equal to 100 nm. These aggregates are formed by structure reproduction replicative mechanism with nonlinear feedback. In this case nanofiller contents discrete change points are bifurcation points. The synergetics methods can be used for prediction of forming nanoparticles aggregates size. The parameters, characterizing interfacial phenomena in polymer nanocomposites (for example, interfacial layer thickness) obey also to synergetics laws.

REFERENCES [1]

[2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]

Mikitaev A.K., Kozlov G.V., Zaikov G.E. Polymer Nanocomposites: Variety of Structural Forms and Applications. New York, Nova Science Publishers, Inc., 2008, 319 p. Kozlov G.V., Mikitaev A.K. Polymers as Natural Nanocomposites: Unrealized Potential. Saarbrücken, Lambert Academic Publishing, 2010, 323 p. Folmanis G.E. Proceedings of Intern. Interdisciplinary Symposium “Fractals and Applied Synergetics”. Moscow, Publishers MSOU, 2003, p. 303-308. Korzhikov A.V., Ivanova V.S. Proceedings of Intern. Interdisciplinary Symposium “Fractals and Applied Synergetics”. Moscow, Publishers MSOU, 2003, p. 278-280. Folmanis G.E. Proceedings of Intern. Interdisciplinary Symposium “Fractals and Applied Synergetics”. Moscow, Publishers MSOU, 2003, p. 284-286. Kozlov G.V., Sultonov N.Zh., Shoranova L.O., Mikitaev A.K. Naukoemkie Tekhnologii, 2011, v. 12, № 6, p. 32-36. Kozlov G.V., Mikitaev A.K. Nanotekhnologii. Nauka i Proizvodstvo, 2011, № 4, p. 57-63. Honeycombe R.W.K. The Plastic Deformation of Metals. Cambridge, Edward Arnold Publishers, Ltd., 1968, 403 p. Balankin A.S. Synergetics of Deformable Body. Moscow, Publishers Ministry of Defence SSSR, 1991, 404 p. Kozlov G.V., Sanditov D.S. Anharmonic Effects and Physical-Mechanical Properties of Polymers. Novosibirsk, Nauka, 1994, 261 p. Kozlov G.V., Zaikov G.E. Structure of the Polymer Amorphous State. Utrecht, Boston, Brill Academic Publishers, 2004, 465 p. Ivanova V.S. Strength and Fracture of Metal Materials. Moscow, Nauka, 1992, 160 p. Kozlov G.V., Burya A.I., Lipatov Yu.S. Mekhanika Kompozitnykh Materialov, 2006, v. 42, № 6, p. 797-802. Hentschel H.G.E., Deutch J.M. Phys. Rev. A, 1984, v. 29, № 12, p. 1609-1611. Wu S. J. Polymer Sci.: Part B: Polymer Phys., 1989, v. 27, № 4, p. 723-741. Aharoni S.M. Macromolecules, 1983, v. 16, № 9, p. 1722-1728. Bobryshev A.N., Kozomazov V.N., Babin L.O., Solomatov V.I. Synergetics of Composite Materials. Lipetsk, NPO ORIUS, 1994, 154 p. Brady L.M., Ball R.C. Nature, 1984, v. 309, № 5965, p. 225-229.

78


[19] Avnir D., Farin D., Pfeifer P. Nature, 1983, v. 308, № 5959, p. 261-263. [20] Kozlov G.V., Yanovskii Yu.G., Zaikov G.E. Structure and Properties of ParticulateFilled Polymer Composites: the Fractal Analysis. New York, Nova Science Publishers, Inc., 2010, 282 p.



EXPERIMENTAL STUDY OF WAX DEPOSITION IN SINGLE-PHASE SUBCOOLED OIL PIPELINES O. A. Adeyanju and L. O. Oyekunle University of Lagos, Nigeria

ABSTRACT The ability to determine the severity of wax deposition is an extremely important issue, particularly in the design and development of deepwater oilfields. Though much progress has been made in the last decades to better the understanding of this complex process, yet the ability to accurately account for all the factors that affect wax deposition are currently not in existence in the wax simulators used presently in the industries. In this study an experimental methodology constructed to simulate wax deposition process was employed to investigate the influence factors controlling paraffin wax deposition to the pipe wall surface (namely, inlet oil temperature, inlet coolant temperature, oil flow rate and the wax content). Series of tests were designed to determine the effects of these influence factors on the wax content in the deposit. The experimental results revealed that the amount of wax deposited initially increases with time, attained a maximum value and gradually erode off. Also it was discovered that the wax deposition decreases with flow rates and also with the temperature difference between the flowing oil and the pipe wall, when the oil temperature is above its Wax Appearance Temperature (WAT), while the reverse is the case when the oil temperature is below its WAT. The study also established that shear dispersion, defined as the movement of wax crystals towards the pipe wall as a result of the velocity variation along the radial direction during oil flow in the pipe ignored in most of the existing models used in the existing wax deposition commercial codes was found not to be inconsequential. The flow rate rather than the flow regime was also discovered to responsible for the shear stripping of wax deposit at the wall. This experimental observation will provide a reference point and an insight for further study on wax deposition in actual pipelines. This is particularly so for oil characterized by high wax content and high gel point temperature like those produced from most fields in Nigeria’s Niger Delta.

NOMENCLATURE d = diameter (m) L = length (m) R = pipeline radius (m) 


80

O. A. Adeyanju and L. O. Oyekunle T = temperature (K)

Greek Letters

 = Fluid density (kg/m3)  = wax thickness (ratio)  = viscosity (cp)

INTRODUCTION Risk associated with the transportation of waxy crudes is one of the most critical operations hazards of deepwater offshore pipelines. The challenge that engineers will face in offshore operation is, thus, how to design the pipeline and subsea system to assure that multiphase waxy crudes will be safely and economically transported from the bottom of the wells through deepwater pipelines all the way to the downstream processing plant (Guo et al, 2005). The practice of identifying, quantifying and mitigating of all the flow risks associated with offshore pipelines and subsea systems is called flow assurance. Flow assurance is tedious for deepwater pipelines and system operations. In deepwater, the seawater temperature is usually much colder than the fluid temperature inside the pipeline (Al-Yaari, 2011). If the fluid temperature inside the pipelines becomes too low due to heat loss, wax deposit will form on the pipe wall. Thus, effective protection of fluid heat is one of the most important design parameters for offshore pipeline. Thus, factors affecting wax deposition mechanisms need to be extensively studied. Wax deposition occurs when paraffinic components in crude oil (alkanes with carbon numbers greater than 20) precipitate and deposit on cold pipeline wall when the inner wall temperature falls below the Wax Appearance Temperature (WAT) (solubility limit). Whereas wax precipitation during oil flow results in wax deposition and flow restriction, wax precipitation during a production shutdown results in problems when attempting to restart the flow ( Al-Yaari, 2011). When the transportation in a pipeline is stopped due to a planned maintenance or an emergency situation such as severe weather conditions on offshore platform the temperature and solubility of wax decreases and wax molecules precipitate out of liquid phase in static condition (Fung et al, 2006, Thomason, 2000) When waxy oil flowing in cold lines is cooled, it gels due to the formation of a network of wax crystals. Unlike in the case of inorganic solutions, where there is hardly any interaction among the salt crystals, the wax crystals have a strong interaction and affinity, resulting in the formation of the network. Although oil (solvent) and wax (solute) have a similar chemical nature, their molecular weights are quite different. Waxes have a higher molecular weight and they tend to form stable wax crystals that interlock to form a solid network. The network of wax traps a large quantity of oil (Holder and Winkler, 1965). Hence, the initial stage of the deposition of the waxy oil mixture on a cold surface is the formation of a gel layer with large fraction of trapped oil.

Experimental Study of Wax Deposition in Single-Phase Subcooled Oil Pipelines The wax deposition (Hernandez et al, 2004).

process

can

be

described

by

the

following

81 steps

1. Gelation of the waxy oil (formation of incipient gel layer) on the cold surface. 2. Diffusion of waxes (hydrocarbon with carbon numbers greater than the critical number) towards the gel layer from the bulk oil. 3. Internal diffusion of these molecules through the trapped oil. 4. Precipitation of these molecules in the deposit. 5. Counter-diffusion of de-waxed oil (hydrocarbon with carbon numbers lower than the critical carbon number) out of the gel layer. The last three steps result in an increase of the solid wax content of the deposit. Though there is ongoing efforts to gain insight into the physical phenomenon of wax deposition, a model to give reliable guesses of the wax build up is still lacking (Gjermundsen and Duenas, 2006; Mehrotra and Bhat, 2007). A reliable wax deposition experimental study will be an invaluable tool for the development of such effective model for optimal scheduling and removal of the deposited gel. Various mechanisms, by which wax deposition could occur, such as molecular diffusion (when the temperature variation in radial direction makes the dissolved wax diffuse from bulk towards the pipe wall), shear dispersion (deposition of the already precipitated wax by shear dispersion), Shear Stripping reduction (deposition rate reduction due to shear stripping) , Brownian diffusion, and gravity settling; have been proposed (Bern et al., 1980; Burger et al. 1981; Majeed et al., 1990). Mechanism such as shear dispersion, Brownian diffusion, and gravity settling had been identified to play a role only for particulate deposition of wax. However, most models studied in the literature neglected the effect of particulate deposition as they believe its effect is inconsequential. Hence, molecular diffusion and shear stripping are currently considered as the predominant mechanisms underlying the wax deposition process. A series of experimental studies had been carried out to study the mechanisms affecting wax deposition, Vankatesan and Creek, (2007) in their study articulated the difference between the laboratory conditions and those prevailing in the fields and recommended that the existing laboratory methods should be reviewed. Hilbert, (2010) experimentally studied the effect of different crude mixtures, emulsion, and pipeline cooling properties on wax behaviours in subsea pipelines and concluded that the shear yield stress of the pipeline fluid decreases with increasing water cut. Al-Yaari, (2011) used a experimental petroleum production model, to review the wax deposition problem in flow and during shut in condition and proposed important wax deposition processes and mechanisms that will enhance the wax deposition study. Dwivedi, et al., (2012) conducted an experimental study in a small-scale loop to determine the effect of turbulence/shear and thermal driving force on wax deposition. It was observed that the paraffin deposition is highly dependent on the thermal effective drive force which is the temperature difference between bulk oil and initial inner pipe wall and also on the turbulence effect. There are currently little studies on wax particulate deposition in the literature hence further study on this area cannot be over emphasized. In this study, the oil a field in Nigeria characterized by high wax content and high gel point temperature was selected as the experimental sample. Different mechanisms leading to wax deposition such as molecular diffusion, shear dispersion and shear removal were

82

O. A. Adeyanju and L. O. Oyekunle

modeled, and key factors affecting wax deposition in pipelines (namely, inlet oil temperature, inlet coolant temperature, oil flow rate and the wax content) were experimentally studied. Preliminary investigations were first conducted on the oil sample to determined its Wax Appearance Temperature (WAT) using a programmable Rheometer that measured the viscosity of the crude at different temperatures under different shear rates, the result is given in figure 1, the WAT of the sample oil was determined to be 43oC, which is the temperature at which the oil start to exhibit non-Newtonian behavior i.e. viscosity starts to show dependence on the shear rate.

EXPERIMENTAL FACILITY DESCRIPTION The experimental setup consisted of test flow loop that is shown in figure 2. This flow loop was used to perform the wax deposition experiments under the single oil phase conditions. This flow loop is made of mild steel pipe of length 140 cm with an inside diameter of 1.5cm. The experimental setup has two sections, the test section and the reference section. The crude oil temperature was regulated with a temperature regulators, the oil is pumped through the test section and the then through reference after passing through the liquid mass flow-meter along the flow lines.

Figure 1. Viscosity profile of the oil sample at different shear rate at different temperature.

The test section is jacketed with a steel jacket in which cold water pumped from a cooling bath are circulated. The purpose of the test section is to maintain the inner pipe wall at a lower temperature than both the bulk oil temperature and Wax Appearance Temperature (WAT) so as to generate the wax deposit on the inner pipe wall, just like what would be encountered in actual pipelines.

Experimental Study of Wax Deposition in Single-Phase Subcooled Oil Pipelines

83

The configuration of the reference section is completely identical with the test section. However, contrary to the test section, the inner pipe wall temperature in the reference section is maintained at a higher temperature than the bulk oil temperature to prevent wax deposition by circulating the heated water into the jacket of reference section. Thermocouples were placed both at the inlet and outlet of the test tube and the reference tube to determine the temperatures at both the inlet and outlet of the test and reference section. Thermocouples were also attached to the cooling water tank and crude oil tank to take temperature reading. The oil in the previous experimental run was ensures to be removed by flowing hot oil for few hours before the commencement of subsequent experimental run.

METHODOLOGY OF THE EXPERIMENT Flow loop experiments were performed to observe the growth and aging of the gel deposit. The waxy oil sample was made to enter the test section at a relatively higher temperature than the wall/coolant temperature in order to generate wax deposit in the inner section of the flow-line. Experiments for four different flow rates 1.0 liter/min, 1.4 liter/min, 1.8 liter/min and 2.2 liters/min, were carried out at the same aging time of about 26 hours to determine the effect of oil temperature on wax deposition. The experiments were performed: 









At the same flow rate of 1.4 liter/min., at the same oil temperature of 55oC, but at different wall temperatures of 40oC, 35oC and 30oC to investigate the dependence of wax deposition on the pipe wall temperature. At the same flow rate of 1.4 liter/min. and at a wall temperature of 37oC at different bulk oil temperatures of 52oC and 56oC to study the effect of oil temperature on wax deposision. At the same flow rate of 1.4 liter/min. and at the wall temperature of 28oC at different oil temperatures of 34oC, 37oC and 40oC which are below the Wax Appearance Temperature (WAT) of the crude to determined the diffusion effect of the precipitated wax. The experiments were then repeated at wall temperature of 30oC, oil temperature of 55oC at four different aging time of 6 hours, 12 hours, 18 hours and 24 hours at a constant flow rate of 1.0 liter/min and samples of the wax-oil gel deposit were collected from the wall of the tubing after each experiment and were analyzed for the wax content through their densities determination. The effect of shear dispersion was investigated by neutralizing the molecular diffusion effect by putting the oil and wall temperature at 38oC which has to be below the Wax Appearance Temperature (WAT) of the crude earlier determined to be 43oC in order to allow the wax to be precipitated and flowing it at a rate of 1.4 liter/min.

84


Figure 2. Schematic diagram of wax deposition test flow loop.

WAX DEPOSITION DETERMINATION METHOD The pressure drop method was used to determine the thickness of the deposited wax. This method is based on the concept that wax deposition in a pipe section reduces the hydraulic diameter of the flowing fluid inside the pipe, resulting in an increase in frictional pressure drop over the pipe section. The pressure drop method is an on-line wax measurement technique that does not require depressurization and restart in order to obtain the measurements. Neither does it impose any influence on the in-situ and overall heat transfer. Once the frictional pressure drop across a pipe section is measured and the flow rate, density and viscosity of the crude oil in the pipe section are determined, the wax thickness present in the pipe wall can be calculated accurately from the following equation (Chen et al, 1997) n

(d i  2 w )

5 n

2cL     4Q       Pf      

2 n

(1)

where Pf is the pressure drop L is the length of pipe section, d is the hydraulic diameter or effective inside diameter, Q is the volumetric flow rate, ρ is the fluid density, Where  is the apparent viscosity of the crude oil. c = 16, n = 1 for laminar flow and c = 0.046, n = 0.2 for turbulent flow. Laminar flow exists when N Re  2000 (Chen et al. ,1997).


85

RESULT AND DISCUSSION Generally it was observed in all the experimental runs that the wax thickness initially increases gradually reaching a peak and then starts diminishing. This is due to wax-oil gel formed initially consisting of oil entrapped in the wax deposit until the oil start diffusing out of the gel deposit at the later period when the wax-gel deposit begins to harden.

EFFECT OF INLET COOLANT TEMPERATURE As in the literature, that wax precipitates out of the transported crude oil to form wax deposit at the wall of cold pipe when oil temperature drop below Wax Appearance Temperature (WAT) of the crude. It has been recognized that the pipe wall temperature has a significant impact on wax deposition. Consequently, the effect of the pipe wall temperature on wax deposition had been studied experimentally and reported in a lots of recently published literature source (Dwivedi, 2012; Semenov, 2012; Noville and Naveira, 2012; Leontaritis and Geroulis, 2011; Al-Yaari, 2011), where the thickness of wax deposits is believe to increase with the decreasing pipe wall temperature (Coolant temperature). Considering the practical situation where the ambient temperature may be around the gel point temperature of the waxy crude oil in subsea pipeline, accordingly, the effect of the pipe wall temperature around the gel point temperature of the oil becomes the research emphasis. Obviously, adjusting the inlet coolant temperature can control the pipe wall temperature in the experiments. The inlet coolant temperatures studied in the experiment were selected at 40oC, 35oC and 30oC respectively, while the inlet oil temperature was kept constant at 55oC. However unlike the convectional phenomenon that the thickness of wax deposit would increase with the decreasing inlet coolant temperature, from the result shown in Figure (3) It can be seen that under the same inlet oil temperature conditions the dimensionless thickness of wax deposit actually decreases with the decreasing inlet coolant temperature around the gel point temperature. The reasonable explanation for the phenomenon is the fact that when the inlet coolant temperature is lowered around the gel point of the oil, the viscosity of the oil at the liquiddeposit interface will increase sharply as the inlet coolant temperature decreases. Figure (1), shows that the oil viscosity at temperature of 35oC is almost 50% higher than that of 40oC. Consequently, the higher oil viscosity close to pipe wall can lead to the following significant influences: 



Under the same oil velocity conditions, the shear stress at the liquid-deposit interface will increase sharply with increasing oil viscosity at the liquid-deposit interface caused by the decreasing inlet coolant temperature, this will enhance the effect of shear stripping, leading to decrease in the amount of wax deposit. In terms of the Fick’s mass fusion law, both the radial temperature gradient and the concentration gradient with respect to temperature will increase as the inlet coolant temperature decreases under the same oil temperature conditions, which will have a positive effect on wax deposition, however, the increasing oil viscosity at the liquid-

86

O. A. Adeyanju and L. O. Oyekunle deposit interface caused by the decreasing inlet coolant temperature will diminish the molecular diffusion coefficient, which will have a negative effect on wax deposition.

Hence, whether the wax deposition rate increases or not depends on which influence factor mentioned is dominant. The combination of the two aspects mentioned above ultimately leads to decrease in the thickness of wax deposit with decreasing inlet coolant temperature. In other words, both the shear stress and the oil viscosity at the liquid-deposit interface are dominant in the process of wax deposition for the inlet coolant temperature being around the gel point temperature of the oil.

EFFECT OF INLET OIL TEMPERATURE In the previous literature (Dwivedi, 2012; Semenov, 2012; Noville and Naveira, 2012), the selection of oil temperature which were studied in the experiment was differentiated in term of wax appearance Temperature (WAT) of the oil sample. However, whether the bulk oil temperature is below its WAT or not does play an important role in the precipitation of wax molecules. Therefore, in the study the selection of oil temperatures was differentiated in term of wax appearance temperature of 43oC.

CASE 1: THE INLET OIL TEMPERATURE IS ABOVE ITS WAT In this case, the inlet oil temperatures studied in the experiment were selected at 48oC, 52 C and 56oC, respectively. The inlet coolant temperature was kept constant at 37oC, Figure (5) shows the dimensionless thickness of the wax deposits as a function of inlet oil temperature at oil flow rate of 1.4 liter/min. The results indicate that the thickness of wax deposits decreases with the increasing inlet oil temperature which is above the WAT under the fixed inlet coolant temperature and oil velocity conditions. The reasons may be due to the fact that when the bulk oil temperature is above its WAT, there are no wax molecules precipitating out of the bulk oil, which results in the decrease in wax deposition with increasing oil temperature. In addition, under the fixed inlet coolant temperature conditions, the higher oil temperature will generate higher temperature at the liquid-deposit interface which can increase the solubility of wax molecules, ultimately leading to the fewer amounts of wax deposits. o

CASE 2: THE INLET OIL TEMPERATURE IS BELOW ITS WAT In this case, the inlet oil temperatures studied in the experiments were selected at 34oC, 37 C and 40oC, respectively. The inlet coolant temperature was kept constant at 28oC, figure (6) show the dimensionless thickness of wax deposit as a function of inlet oil temperature for different oil velocities. Contrary to the results in case 1, the results of case 2 indicate that the o


87

thickness of wax deposits increases with increasing inlet oil temperature which is below the WAT under the fixed inlet coolant temperature and oil velocity conditions. The reasons may be due to: 



When the coolant temperature is below its WAT, the zone of wax precipitation will be enlarged with the increasing oil temperature accordingly, enlarged zone of wax precipitation caused by the increase in oil temperature will be prone to make more amount of wax deposit. For the fixed inlet coolant temperature, the higher inlet oil temperature can exert much bigger thermal driving force (i.e. temperature difference) which enhances its potential to generate more amounts of wax deposits.

EFFECT OF OIL FLOW RATE Flow regimes were generally believe to have great impacts on wax deposition (i.e. the higher oil flow rate could generate more wax deposit for the laminar flow regime, and less wax deposit for the turbulent flow regime due to the effect of shear stripping). In this study, the oil flow rate rather than flow regime is considered the dominant factor affecting the thickness of wax deposit. In the laminar flow conditions the wax deposit was observed to be govern by the following principles;  The increased oil flow rate can increases the shear stress at the liquid-deposit interface, which will reinforce the intensity of shear stripping, consequently reducing the wax deposit thickness.  The internal heat transfer coefficient increases with the increasing oil flow rate for laminar flow, which increases the radial temperature gradient, consequently leading to increase in deposit wax thickness. Obviously, these two effects caused by oil velocity are simultaneous. Hence, whether wax deposition is promoted or hindered depends on which one of the two effects caused by oil velocity/flow rate is dominant in the process of wax deposition. In the experimental study, as the experiment were performed in both laminar and turbulent flow regimes, the results show that as the velocity increases from laminar flow regime (flow rates of 1.00, 1.40, and 1.80 liter/min) to turbulent flow regime (flow rate of 2.20 liter/min), the effect of shear stripping becomes dominant FigURE (6), with the deposit thickness increasing when the radial temperature gradient is dominant (i.e. during the laminar flow regime) and deposit thickness decreases when the shear stripping is dominant (during the turbulent flow regime). The wax that deposit at a higher flow rate is harder and more compact judging by the increase in density of the deposited wax during the turbulent flow regime. In other words, only those wax crystals and crystal cluster capable of firm attachment to the surface, with good cohesion among themselves will not be removed from the deposit surface (Kelechukwu et al, 2010). In conclusion, it is not the flow regime but the fluid velocity that is responsible for shear stripping.

88


Figure 3. Dimensionless thickness of wax deposit versus time at different wall temperatures at oil flow rate of 1.4 liter/minutes.

Figure 4. Dimensionless thickness of wax deposit versus time at different inlet oil temperatures above WAT at inlet coolant temperature of 37oC and oil flow rate of 1.4 liter/min.


89

Figure 5. Dimensionless thickness of wax deposit versus time at different inlet oil temperatures below WAT at inlet coolant temperature of 28oC (below WAT) at oil flow rate of 1.4 liter/min.

Figure 6. Dimensionless thickness of wax deposit versus time under different flow rates at oil temperature of 56oC at wall temperature of 37oC.

SENSITIVITY INVESTIGATION OF VARIOUS SHEAR DISPERSION INFLUENT FACTORS The effects of various factors influencing the wax deposition tendency during shear dispersion/gravity settling processes, which include bulk wall and oil temperature, the oil flow rate and the wax content of the oil sample, were experimentally investigated in term of

90


dimensionless wax deposit defined as the ratio of the wax thickness to the initial inner radius of the pipe,

 R

,in order to determine their effect on wax deposit.

EFFECT OF BULK OIL AND WALL TEMPERATURE DEPRESSION BELOW THE WAT Keeping the wax content at 20% and oil flow rate at 1.4 liters/min. and varying the bulk oil and wall temperature showed that the amount of wax deposition increases with decrease in bulk oil and wall temperature (as shown in Figure 7). This is due to the facts that more wax precipitates as the temperature depress further below the wax appearance temperature, leading to bigger wax crystal formation in the bulk oil. Bigger the wax crystal indicate higher the shear dispersion coefficient, this is in agreement with the Bhattacharya (1991) equation.

EFFECT OF FLOW RATE Decreasing the flow rates in the experimental run and keeping the bulk oil and wall temperature at 40oC and wax content at 20%, result in increase in the dimensionless wax deposit thickness after twenty-eight (28) hours of experimental run. This is partly due to the increase in velocity gradient in the radial direction as a result of decrease flow rate and partly due to the longer residence time of the waxy oil in the pipe-line, resulting in more wax crystals diffusing towards the pipe wall. This is shown in Figure 8.

Figure 7. Plot of Dimensionless wax thickness due shear dispersion effect against time at different temperatures.


91

EFFECT OF WAX CONTENT Increasing the wax content of the sample oil to 32% by adding the wax deposit scrapped from earlier experimental run to a new oil sample results in the increase in the wax deposit thickness increasing after almost twenty-eight (28) hours of experimental run.

Figure 8. Plot of Dimensionless wax thickness due shear dispersion effect against time at different flow rates.

Figure 9. Plot of Dimensionless wax thickness due shear dispersion effect against time at different wax contents.

This may be due to more wax crystals available for diffusion towards the pipe wall at that temperature below the WAT as the wax content in the oil increases; this ultimately leads to increase in wax thickness due to increase in shear dispersion/gravity settling.

92


Comparison of Shear Dispersion and Molecular Diffusion Contribution to Wax Deposition Processes Figure 10 shows the comparison between the deposited wax due to shear dispersion and those due to molecular diffusion, though more wax is deposited through the molecular diffusion process, the effect of shear dispersion on wax deposition cannot be assume to be inconsequential as assumed by most of the current wax deposition models. Hence the wax deposition is more of temperature gradient than velocity gradient process.

Figure 10. Plot of Dimensionless wax thickness due shear dispersion and molecular diffusion effect against time.

The Wax Content of the Deposit Changes with Time The change in wax fraction in the deposited wax was confirmed by changes in the density of deposited wax with time as shown in figure 11. This agrees with the theoretical concept of wax continuously diffusing into the deposit and the oil seeping out of the deposit ultimately leading to the net wax deposit density to increase with time.

Figure 11. The change in density of the deposited wax deposit with time at oil flow rate of 1.00 liters/min.


93

OUTLET TEMPERATURE OF TEST SECTION VARIATION WITH TIME At initial oil temperature of 55oC and coolant temperature of 35oC, the change in outlet bulk oil and coolant temperatures was observed at different periods. The result is as shown in Figure 9, where the outlet bulk oil temperature increases with time, while the outlet coolant temperature decreases with time. This indicates that the wax deposit acts as an insulator, thereby reducing the heat lost by the bulk oil to the pipe wall as the wax deposit grows.

Figure 12. Outlet temperatures of the bulk oil and the water/wall temperature at different time.

CONCLUSION The conclusion established based on the observation, discussions, and applications carried out in this study is summarized as the followings 





Contrary to the general theory that wax deposit increases with an increase in the temperature difference between the pipe wall and the bulk oil, it was observes that when the oil temperature is above its Wax Appearance Temperature (WAT) and the pipe wall temperature below its WAT, the wax deposit decreases with an increase in temperature difference, while the reverse is the case when both the oil and pipe wall temperature are below the oil WAT. It takes some time before wax start to deposits in a cold pipe during warm crude oil passage through the pipe. As the deposits grow it act as an insulator and reduces the heat exchange between the oil and the pipe wall. The effect of shear dispersion/ gravity settling contributions to wax deposition though relatively less than the contribution by molecular diffusion but it is not negligible as assumed by most of the currently available wax deposition models.

94

O. A. Adeyanju and L. O. Oyekunle 

The shear removal flux in the existing models is currently considered to be constant in time, More dynamic tests, such as those conducted with the crude oil , are required to obtain more information on the shear stripping term. This term needs to be studied as a function of shear stress, Reynolds number, and fluid viscosity.

REFERENCES Al-Yaari, M., (2011) “Paraffin Wax Deposition: Mitigation and Removal Techniques” Presented at the 2011 SPE professionals Technical Symposium held in Dhahran, SaudiArabia. Bhattacharya, A. (1991), “Effect of Thermal Environment on Wax Deposition in a Crude Oil Pipeline: Analysis of some Critical Aspects” Proceedings of the first (1991) International offshore and Polar Engineering Conference, Edinburgh, United Kingdom. Bern, P. A., Withers, V.R. and Cairns, J.R.,(1980), “Wax Deposition in Crude-oil pipelines” Euro. Proc. Offshore Petrol. Conference and Exhibition, London, 571. Burger , E.D., Perkins, T.K. and Striegler, J.H. (1981), “Studies of Wax Deposition in the Tran Alaska Pipeline,” J. Pet. Tech, 33, 1075. Chen, X.T., Volk, B. M., Bwith Stuck Pig,” rill, J.P. (1997), “Techniques for Measuring Wax Thickness during Single and Multi-phase Flow” Presented at SPE Annual Technical Conference and Exhibition held in San Antonio, Texas, USA. Dwivedi, P., Sarica, C. and Chang, W., (2012), “Experimental Study of Wax deposition Characteristics of a Waxy Crude oil under Single Phase Turbulent Flow Conditions” Paper no., OTC 22953, Presented at the 2011 Offshore Technology Conference, Houston, Texas, USA. Fung, G., Bechhaus, W.P., McDaniel, S. and Erdogmus, M.(2006), “To Pig or Not To pig: The Marlin Experience with Stuck pipe” Offshore Technology Conference, Houston, TX. Gjermundsen, I. and Duenas Dietz, M.(2006), In proceedings of the 7th International Conference on Petroleum Phase Behaviour and Fouling; Kilpatrick, Ed., Asheville, NC, USA Guo, B, Song, S, Chacko, J and Ghalambor, A, (2005) “’Offshore Pipelines” Elsivier Inc. Oxford, UK, . Hernandez, O.C., Hensly, H. and Serica, C., (2004), “Improvements in Single-Phase paraffin Deposition Modeling” SPE Production and facilities (11): 237-244. Hilbert, J., (2010), “Flow Assurance: Wax Deposition and Gelling in Subsea Oil Pipelines” SPE 133948, Presented at the 2010 SPE Annual Technical Conference and Exhibition held in Brisbane, Queensland, Australia. Holder, G.A. and Winkler, J.(1965), “Wax Crystallization from Distillate Fuels: I. Cloud and Pour Phenomena Exhibited by Solutions of Binary n-Paraffin Mixtures,. Journal of Institute of petroleum, 51, 235. Kelechukwu, E.M., Al Salim, H.S. and Yassin, A.M. “Influencing Factors governing Paraffin Wax Deposition during Crude Production” International Journal of the Physical Sciences Vol 5(15), 2351-2362. Leontaritis, K. J. and Gerouulis, E., (2011), “Wax Deposition Correlation-Application in Multiphase Wax Deposition Models” Paper no. OTC 21623, Presented at the 2011 Offshore Technology Conference, Houston, Texas, USA.


95

Majeed, A., Bringedal, B. and Overa, S., (1990), “Model to Calculates Wax deposition for N. Sea Oils,” oil Gas J., 88, 63. Mehrotra, A.K., Bhat, N.V., (2007), “Modeling the Effect of Shear Stress on Deposition from Waxy Mixtures under Laminar Flow with Heat Transfer” Energy and Fuels, 21, 1277. Noville, I. and Noveira, L., (2012), “Comparison between Real Field data and the Results of Wax Deposition Simulation” SPE 152575, Presented at the 2012 SPE Latin American and Caribbean petroleum Engineering Conference held in Mexico City, Mexico. Semenov, A., (2012), “Wax Deposition Forecast” SPE 149793, Presented at the 2012 SPE North Africa Annual Technical Conference and Exhibition held in Cairo, Egypt. Thomason, W.H., (2000) “Start-Up and Shut-In Issues for subsea Production of High Paraffinic Crudes” Offshore Technology Conference, Houston, TX. Venkatesan, R., and Creek, J.L.,(2007), “ Wax Deposition during production Operations: SOTA” paper OTC 18798, Presented at the 2007 Offshore Technology Conference, Houston, Texas, USA.



EFFECT OF THE MODIFICATION BY ORGANIC SILICON SUBSTANCES OF THE MINERAL FILLERS ON SOME PROPERTIES OF EPOXY RESIN COMPOSITES J. Aneli1, O. Mukbaniani2, E. Markarashvili2, G. Zaikov3, and E. Klodzinska4 1

2

R. Dvali Institute of Machine Mechanics I. Javakhishvili Tbilisi State University Faculty of Exact and Natural Sciences, Department of Macromolecular Chemistry, Tbilisi, Republic of Georgia 3 N. M. Emanuel Institute of Biochemical Physics, Moscow, Russia 4 Institute for Engineering of Polymer Materials and Dyes, Torun, Poland

ABSTRACT Ultimate strength, softening temperature, and water absorption of the polymer composites based on epoxy resin (type ED-20) with unmodified and/or modified by tetraethoxysilane (TEOS) mineral diatomite are described. Comparison of experimental results obtained for investigated composites shows that ones containing modified filler have the better technical parameters mentioned above than composites with unmodified filler at corresponding loading. Experimentally is shown that the composites containing binary fillers diatomite and andesite at definite ratio of them possess the optimal characteristics – so called synergistic effect. Experimental results are explained in terms of structural peculiarities of polymer composites.

Keywords: polymer composite, epoxy resin, modified filler, ultimate strength, softening temperature, water absorption, synergistic effect of fillers

1. INTRODUCTION In recent time the mineral fillers attract attention as active filling agents in polymer composites [1, 2]. Thanks to these fillers many properties of the composites are improved increases the durability and rigidity, decrease the shrinkage during hardening process and water absorption, improves thermal stability, fire proof and dielectric properties and finally 

[email protected]

98

J. Aneli, O. Mukbaniani, E. Markarashvili et al.

the price of composites becomes cheaper [3-5]. At the same time it must be noted that the mineral fillers at high content lead to some impair of different physical properties of composites. Therefore the attention of the scientists is attracted to substances, which would be remove mentioned leaks. It is known that silicon organic substances (both low and high molecular) reveal hydrophobic properties, high elasticity and durability in wide range of filling and temperatures [6, 7]. The purpose of presented work is the investigation of effect of modify by TEOS of the mineral –diatomite as main filler and same mineral with andesite (binary filler) on some physical properties of composites based on epoxy resin.

2. BASIC PART Mineral diatomite as a filler was used. The organic solvents were purified by drying and distillation. The purity of starting compounds was controlled by an LKhM-8-MD gas liquid chromatography; phase SKTF-100 (10%, the NAW chromosorb, carrier gas He, 2m colomn). FTIR spectra were recorded on a Jasco FTIR-4200 device. The silanization reaction of diatomite surface with TEOS was carried out by means of three-necked flask supplied with mechanical mixer, thermometer and dropping funnel. For obtaining of modified by 3 mass % diatomite to a solution of 50 g grind finely diatomite in 80 ml anhydrous toluene the toluene solution of 1.5 g (0.0072 mole) TEOS in 5 ml toluene was added. The reaction mixture was heated at the boiling temperature of used solvent toluene. Than the solid reaction product was filtrated, the solvents (toluene and ethyl alcohol) were eliminated and the reaction product was dried up to constant mass in vacuum. Other product modified by 5% tetraethoxysilane was produced via the same method. Following parameters were defined for obtained composites: ultimate strength (on the stretching apparatus of type “Instron”), softening temperature (Vica method), density and water absorption (at saving of the corresponding standards).

RESULTS AND DISCUSSION High temperature condensation reaction between diatomite and TEOS from the one side and between andesite and same modifier from the other one was carried out in toluene solution (~38%). The masses of TEOS were 3 and 5% from the mass of filler. The reaction systems were heated at the solvent boiling temperature (~110oC) during 5-6 hours by stirring. The reaction proceeds according to the following scheme:

OH OH + Si(OC2H5)4 OH

O-Si(OC2H5)3 T 0C -C2H5OH

O-Si(OC2H5)2-O OH

Effect of the Modification by Organic Silicon Substances of the Mineral Fillers …

99

The direction of reaction defined by FTIR spectra analysis shown that after reaction between mineral surface hydroxyl, -OSi(OEt)3 and the -OSi(OEt)2O- groups are formed on the mineral particles surface. In the FTIR spectra of modified diatomite one can observe absorption bands characteristic for asymmetric valence oscillation for linear Si-O-Si bonds at 1030 cm-1. In the spectra one can see absorption bands characteristic for valence oscillation of Si-O-C bonds at 1150 cm-1 and for C-H bonds at 2950-3000 cm-1. One can see also broadened absorption bands characteristic for unassociated hydroxyl groups. On the basis of modified diatomite and epoxy resin (of type ED-20) the polymer composites with different content of filler were obtained after careful wet mixing of components in mixer. After the blends with hardening agent (polyethylene-polyamine) were placed to the cylindrical forms (in accordance with standards ISO) for hardening, at room temperature, during 24 h. The samples hardened later were exposed to temperature treatment at 120oC during 4 h. The concentration of powder diatomite (average diameter up to 50 micron) was changed in the range 10-60 mass %. The curves on the Figure 1 show that at increasing of filler (diatomite) concentration in the composites the density of materials essentially depends on both of diatomite contain and on the degree of concentration of modify agent (TEOS). Naturally the decreasing of density of the concentration of unmodified (1), modified by 3% (2) and 5 mass % (3) tetraethoxysilane diatomite The dependence of ultimate strength on the content of diatomite (modified and unmodified) presented on the Figure 2 shows that it has an extreme character. However the positions of corresponding curves maximums essentially depend on amount of modified agent TEOS. The general view of these dependences is in full conformity with well known dependence of σ – C [8].

Figure 1. Dependence of the density of the composites based on epoxy resin on composites at increasing of filler concentration is due to increasing of micro empties because of one’s localized in the filler particles (Figure 1, curve 1). The composites with modified by TEOS diatomite contain less amount of empties as they are filled with modify agent (Figure 1, curves 2 and 3).

100


Figure 2. Dependence of ultimate strength of the composites based on ED-20 with unmodified (1) and modified by 3 (2), and 5 mass % (3) TEOS diatomite.

The sharing of the maximum of curve for composites containing 5% of modified diatomite from the maximum for the analogous composites containing 3% modifier to some extent is due to increasing of the amount of the bonds between filler particles and macromolecules at increasing of the concentration of the filler. Investigation of composites softening temperature was carried out by apparatus of Vica method. Figure 3 shows the temperature dependence of the indentor deepening to the mass of the sample for composites with fixed (20 mass %) concentration of unmodified and modified by TEOS.

Figure 3. Temperature dependence of the indentor deepening in the sample for composites containing 0 (1), 20 mass % (2), 20 mass % modified by 3% TEOS (3), 20 mass % modified by 5% TEOS (4) diatomite.

Based on character of curves on the Figure 3 it may be proposed that the composites containing diatomite modified by TEOS possesses thermo-stability higher than in case of analogous composites with unmodified filler. Probably the presence of increased interactions between macromolecules and filler particles due to modify agent leads to increasing of thermo-stability of composites with modified diatomite.

Effect of the Modification by Organic Silicon Substances of the Mineral Fillers … 101 Effect of silane modifier on the investigated polymer composites reveals also in the water absorption. In accordance with Figure 4 this parameter is increased at increasing of filler contain. However if the composites contain the diatomite modified by TEOS this dependence becomes weak.

Figure 4. Dependence of the water-absorption on the concentration of filler in the composites based on epoxy resin containing diatomite modified by 5% (1) and 3% (2) tetraethoxysilane and unmodified (3) one.

Figure 5. Dependence of the density on the concentration of diatomite in binary fillers with andesite. (1) - unmodified and modified by 5% tetraethoxysilane (2) fillers for composites based on epoxy resin. Full concentration of binary filler in composites 50 mass %.

There were conducted the investigation of binary fillers on the properties of the composites with same polymer basis (ED-20). Two types of minerals diatomite and andesite with different ratios were used as fillers. It was interesting to establish effect both of ratio of the fillers and effect of modifier TEOS on the same properties of the polymer composites investigated above. The curves presented on the Figure 5 show the effect of modify agent TEOS on the dependence of the density of composites containing the binary filler diatomite and andesite on

102


ratio of lasts when the total content of fillers is 50 mass % to which the maximal ultimate strength corresponds. The maximum of noted effect corresponds to composite, filler ratio diatomite/andesite in which is about 20/30. Probably microstructure of such composite corresponds to optimal distribution of filler particles in the polymer matrix at minimal inner energy of statistical equilibration, at which the concentration of empties is minimal because of dense disposition of the composite components. It is known that such structures consists minimal amount both of micro and macro structural defects [8]. Such approach to microstructure of composites with optimal ratio of the composite ingredients allows supposing that these composites would be possessed high mechanical properties, thermo-stability and low water-absorption. Moreover the composites with same concentrations of the fillers modified by TEOS possess all the noted above properties better than ones for composites with unmodified by TEOS binary fillers, which may be proposed early (Figures 6-8). Indeed the curves on the Figures 6-8 show that the maximal ultimate strength, thermo-stability and simultaneously hydrophobicity correspond to composites with same ratio of fillers to which the maximal density corresponds. The obtained experimental results may be explained in terms of composite structure peculiarities. Silane molecules displaced on the surface of diatomite and andesite particles lead to activation of them and participate in chemical reactions between active groups of TEOS (hydroxyl) and homopolymer (epoxy group). Silane molecules create the “buffer” zones between filler and the homopolymer. This phenomenon may be one of the reasons of increasing of strengthening of composites in comparison with composites containing unmodified fillers. The composites with modified diatomite display more high compatibility of the components than in case of same composites with unmodified filler.

Figure 6. Dependence of the ultimate strength on the concentration of diatomite in binary fillers with andesite. (1) - unmodified fillers and modified by 5% tetraethoxysilane (2) ones for composites based on epoxy resin Full concentration of binary filler in composites 50 mass %.

The modified filler has more strong contact with polymer matrix (thanks to silane modifier) than unmodified diatomite. Therefore mechanical stresses formed in composites by stretching or compressing forces absorb effectively by relatively soft silane phases, i.e. the

Effect of the Modification by Organic Silicon Substances of the Mineral Fillers … 103 development of micro defects in carbon chain polymer matrix of composite districts and finishes in silane part of material the rigidity of which decreases.

Figure 7. Thermo-stability of composites with binary fillers at ratio diatomite/andesite =20/30.

Figure 8. Dependence of the water-absorption of composites based on epoxy resin on the concentration of diatomite in binary fillers with andesite. (1) - unmodified and modified by 5% tetraethoxysilane (2) fillers. Total concentration of binary fillers in composites 50 mass %.

The structural peculiarities of composites display also in thermo-mechanical properties of the materials. It is clear that softening of composites with modified by TEOS composites begins at relatively high temperatures. This phenomenon is in good correlation with corresponding composite mechanical strength. Of course the modified filler has more strong interactions (thanks to modifier) with epoxy polymer molecules, than unmodified filler.

104


The amplified competition of the filler particles with macromolecules by TEOS displays well also on the characteristics of water absorption. In general loosening of micro-structure because of micro empty areas is due to the increasing of filler content. Formation of such defects in the microstructure of composite promotes the water absorption processes. Water absorption of composites with modified diatomite is lower than that for one with unmodified filler to some extent. The decreasing of water absorption of composites containing silane compound is result of hydrophobic properties of ones. Composites with binary fillers possess so called synergistic effect- non-additive increasing of technical characteristics of composites at containing of fillers with definite ratio of them, which is due to creation of the dens distribution of ingredients in composites.

CONCLUSION Comparison of the density, ultimate strength, softening temperature and water absorption for polymer composites based on epoxy resin and unmodified and modified by tetraethoxysilane mineral fillers diatomite and andesite leads to conclusion that modify agent stipulates the formation of heterogeneous structures with higher compatibility of ingredients and consequently to enhancing of noted above technical characteristics.

REFERENCES [1] [2] [3] [4] [5] [6]

[7] [8]

Katz H.S. Milevski J.V. Handbook of Fillers for Plastics, RAPRA, 1987. Mareri P., Bastrole S., Broda N., Crespi A. //Composites Science and Technology, 1998, 58(5), pp.747 -755. Tolonen H., Sjolind S. //Mechanics of composite materials, 1996, 31(4), pp.317-322. Rothon S.: Particulate filled polymer composites, RAPRA, N-Y, 2003, -205 p. Lou J., Harinath V. // Journal of Materials Processing Technology. 2004, 152(2), pp.185-193. Khananashvili L.M., Mukbaniani O.V., Zaikov G.E. Monograph, New Concepts in Polymer Science, «Elementorganic Monomers: Technology, Properties, Applications». Printed in Netherlands, ///VSP///, Utrecht, (2006). Aneli J.N., Khananashvili L.M., Zaikov G.E. Structuring and conductivity of polymer composites. Nova Sci. Publ., New-York, 1998. -326 p. Zelenev Y.V., Bartenev G.M. Physics of Polymers. M.Visshaya Shkola,1978. -432 p. (in Russian).



SURFACE ACTIVATION OF FIBROUS PET MATERIALS N. P. Prorokova, A. V. Chorev, S. M. Kuzmin, S. Yu. Vavilova, and V. N. Prorokov G. A. Krestov Institute of Solution Chemistry of Russian Academy of Scienses (ISC RAS) Ivanovo, Russia

ABSTRACT In this paper the authors demonstrate the possibility of surface activation of polyethylene terephthalate materials due to their weak alkaline hydrolysis, does not cause weakening of the fibrous material. As a result of hydrolysis of the surface layer of fibrous material formed an additional amount of carboxyl and hydroxyl groups. Last exert a positive influence on application functional products to a surface of polyester material. We also show that in a weak surface hydrolysis initially smooth surface of PET becomes nanoroughness that promotes adhesion of functional products for polyester material. A comparison of the effectiveness of the activating action of aqueous solutions of sodium hydroxide, it’s very dilute solutions with the addition of preparation based on quaternary ammonium compounds and aqueous solutions of ammonia and carbamide. It is established that the largest number of hydroxyl groups, with saving the strength of the fibrous material at the initial level is formed in the processing of PET material with a carbamide solution. It is established that the chemical activation with a carbamide solution of PET fabric, and it undergoing anti-microbial finishing with Sanitized T99-19 and subsequent washing, provides suppression to the growth of pathogenic microorganisms on the fabric. It is shown that the result achieved due to the presence on the surface of activated PET material chemically active groups, on which the molecule of Sanitized T99-19 are fixed and formed an ordered structure.

Keywords: polyester materials; alkaline hydrolysis; chemical activation; fabric antimicrobial finish

1. INTRODUCTION One of the promising ways to create textiles with special consumer features is the formation of filaments on the surface of ultra-thin layer of functional products. For their strong fixation on the fibrous material it must be installed on the surface of chemically active 

[email protected]

106

N. P. Prorokova, A. V. Chorev, S. M. Kuzmin et al.

groups. These groups have fibers based on PET is practically absent (there is only a very small number of terminal hydroxyls and carboxyls). However, it is known that PET in the presence of alkali metal hydroxides, which act as catalysts for ester hydrolysis reactions and the reaction of alkaline hydrolysis initially occurs in the outer area of the fiber 1-3. As a result, on the surface of the PET material the hydroxyl and carboxyl groups are formed (Figure 1).

CH2

COO

COO

CH2CH2O

CH2

COO

COOH + OH

[Na+] , [OH-]

CH2CH2O

Figure 1. Alkaline hydrolysis reaction of PET material.

Alkaline hydrolysis of PET material has been popular in the finishing treatment of the textile industry during 1980-1990 and was used to make synthetic materials silk-like properties. However, the literature indicates that the positive effect due to hydrolysis of the polymer (superior neck, high hydrophilicity, low electrified, improved colorability), is achieved only when the mass loss of fibrous material at least 10 - 30% [1,2]. The result is reduced fiber tensile strength [4]. Significant weight loss is due to the fact that the improvement in the fiber is largely determined by its morphology. Micro relief of surface of polyester fibers subjected to the action of sodium hydroxide, is inherent in the presence of etch pits, the size and number of which are determined by the intensity of treatment [1,5]. We suggest that the formation of active chemical groups on the surface of PET fibers can be achieved under milder conditions of hydrolysis.

2. PET FIBER MODIFICATION BY TREATMENT IN SODIUM HYDROXIDE SOLUTIONS WITHOUT AND WITH QUATERNARY AMMONIUM SALTS ADDITIONS The PET fiber was treated with aqueous solutions of sodium hydroxide (concentration in range 0.0125 - 1.5 mol/l) at the boiling point. The process duration was varied from 5 to 20 minutes. The qualitative observation and identification of new surface groups were recorded using the infrared spectra (ATR-FTIR method). Quantitative determination of localized surface active groups was conducted by two methods. Titrimetric method of the surface carboxyl groups calculating was described elsewhere previously [6]. The method is based on the interaction of the carboxyl groups with calcium acetate. The exchange reaction lied to free acetic acid formation which can be titrated. For

Surface Activation of Fibrous PET Materials

107

titration was used titrated against 0.01 M sodium hydroxide solution in the presence of a mixed indicator consisting of thymol blue and cresol red. It is necessary to note that due to the low content of carboxyl groups on the fibrous material surface a very small amount of acetic acid was formed. Therewith the acetic acid is generated in the presence of large amounts of water so it is strongly diluted. As a result of low sensitivity of the titrimetric method, it was applied only for the approximate estimation of the quantitative content of carboxyl groups on the surface. The colorimetric method for determination of the surface hydroxyl groups’ concentration was developed for PET films [7]. We adapted this method for the fiber material surface study. The method is based on the ability of active dichlortriazine dyes to form a covalent bond with the hydroxyl group of polyethylene terephthalate. The coloring active dichlortriazine bright blue azoic dye was used. Dyeing of samples was performed by the standard technology for active dyes with an index of «azoic dye» [8]. Amount of dye fixed on the fiber was determined by the difference in the color intensity of the modified and unmodified PET fibers. The color intensity was estimated from the color characteristics of the samples of coated polyester material, which was determined using color measurement complex, equipped with the program «Colorist» (version 4.2.1994, '99, the authors Pobedinsky V.S., Telegin F.U., Danilin, I.A.). Strength was assessed by the treated yarn breaking load, as measured using the tensile machine TM-3-1, in accordance with GOST 6611.2-73. Figure 2 shows fragments of the ATR-FTIR spectra of PET films, untreated PET (1) and another treated with sodium hydroxide (2). The appearance of a band at 3200-3500 cm-1 indicates that treated surface layer of polymeric material has a significant additional number of OH groups [9], among of which may be hydroxyls, are part of the carboxyl groups. The change of carboxyl groups of which are localized on the surface of PET material was determined by titrimetric method. As a result of treatment process, the concentration of carboxyl groups was increased, i.e. 2.6 times after treatment in sodium hydroxide solution (concentration 0.375 mol/l) at the boiling temperature during 15 min (from 4·10-4 mol/kg for raw polyester material to 10.4·10-4 mol/kg).

Figure 2. Vibrational spectra of polyester film (range 2700 - 3600 cm-1):1 – untreated (initial) film; 2 – film after treatment by boiling in 0.0125 NaOH solution for 20 min.

108


The new carboxyl groups are localized on the PET surface and are available for interaction. Therefore, the specific surface area was measured independently to characterize the surface carboxyl group density. The calculated values of the surface active groups concentration were around 1.02·10-6 mol/m2 for the initial PET material and 2.66·10-6 mol/m2 after modification. The number of hydroxyl groups formed on the surface of the PET material during the treatment process was determined by the colorimetric method (Figure 3). As follows from the data presented in Figure 3 the processing of PET fiber material with sodium hydroxide solution leads to an increase in concentration of surface hydroxyl groups. The results obtained from the IR spectra of the PET (Figure 2) indicate on the surface localization of a noticeable amount of hydroxyl groups.

Figure 3. The concentration of surface OH groups on the PET fibrous material after treatment. Treatment process using sodium hydroxide concentrations of: 1 - 0.0125 mol/l, 2 - 0.025 mol/l, 3 0.0375 mol/l, 4 – 0.125 mol/l; 5 – 0.25 mol/l; 6 – 0.375 mol/l; 7 – 0.5 mol/l; 8 – 1 mol/l; 9 – 1.5 mol/l.

By comparing the concentrations of surface hydroxyl and carboxyl groups derived from the treatment process, it could be concluded that they are nearly the same. For example, after treatment of the initial PET material by 0.375 M Noah solution for 15 min., the surface hydroxyl and carboxyl concentrations are increased to 2.79·10-6 mol/m2 and 2.66·10-6 mol/m2, respectively. This conformity corroborates the mechanism of hydrolysis, resulting in the formation of the equal number of hydroxyl and carboxyl groups (as shown in Figure 1). Since the colorimetric method allows analyzing several samples at the same time it was used in serial experiments. It should be noted that, in current experiment the Noah concentration differs more than hundred times, but the most significant difference between the numbers of active groups formed is only ~ 40%. The maximum value was recorded during 10-15 min. of treatment


109

process. Processing of PET material with 0.125 M Noah solution leads to the greater number of the surface hydroxyl groups comparing to that of more or less concentrations. Table 1.

Concentration of surface OH groups, mol/m2

NaOH concentration, mol/l

Table 1. Concentration of surface OH groups after treatment process using different concentrations of NaOH for 15 min. and boiling condition

0.0125

2.2310-6

0.025

2.310-6

0.0375

0.125

0.25

0.375

0.5

1

1.5

2.3410-6

3.0710-6

2.9110-6

2.810-6

2.610-6

2.6710-6

2.3310-6

Considering the available literature data on the mechanism and kinetics of alkaline hydrolysis of PET fiber materials [1-3], we assume that if the NaOH concentration more than 0.1 mol/l, the hydrolysis is intensive enough for formation of appreciable amounts of hydroxyl and carboxyl groups on the fiber surface. In contrast, when NaOH concentration is above 0.5 mol/l the so-called "etching" of the fibrous material surface was appeared. It is rapid and intense degradation of the surface layers of polymer material with the formation of low molecular weight hydrolysis products (oligomers and terephthalic acid, ethylene glycol), passing into solution. Thus, most of the reactive functional groups are removed from the surface. Perhaps in a higher concentration they are present in the field of micro-cracks and defects of the fiber, where the deeper penetration of the reagent, accompanied by the formation of micro cavities. The etching of the fibers is most inevitably leading to a decrease in the thickness of the fiber and, consequently surface, to reducing the breaking load. This is confirmed by the data given in Table 2. Presented data in the table indicates a decrease in the strength of the PET material with increasing duration time of treatment and the concentration of sodium hydroxide solution. In the most corrosive environment (concentration of sodium hydroxide - 1.5 mol/l) after 20 min. the tensile strength of PET material is reduced by 15%. The treatment of PET material used 0.125 - 0.375 M NaOH the breaking load of yarn is almost at baseline. Changes in weight of polyethylene fibers in a process of treatment are shown in Table 3. From the data in the table could be stated that the maximum weight loss of PET fiber material treated with a solution of sodium hydroxide concentration of 0.375 mol/l for 20 min. recorded only 1.2%. Therefore, it could be concluded that there is no strong degradation of PET fibers in these conditions.

110


By summarizing the above data, it should state that after treatment of PET fiber material using 0.25 - 0.375 M sodium hydroxide solutions for 10 - 15 min. on the surface new functional groups were appeared. The tensile strength of fibrous material remains practically unchanged. Table 2. The breaking load (cN) of the PET yarn after treatment by boiling in NaOH solutions Treatment time, min

10 15 Concentration, mol/l 0.0125 1200±78 1183±63 0.0250 1206±86 1210±77 0.0375 1161±69 1164±82 0.1250 1173±87 1239±111 0.2500 1122±75 1142±110 0.3750 1134±86 1128±116 0.5000 1089±152 1107±83 1.0000 1061±87 1074±101 1.5000 1060±89 1045±104 Breaking load of untreated yarn 1180±115 centi-Newton (cN)

20 1194±93 1193±81 1159±90 1135±100 1151±103 1130±74 1094±85 1016±70 1000±78

Table 3. The weight loss (%) of the PET yarn after treatment by boiling in NaOH solutions Treatment time, min 10 15 20

0.125 mol/l 0.18 0.44 0.28

0.25 mol/l 0.4 0.82 0.88

0.375 mol/l 0.85 1.20 1.20

0.5 mol/l 0.91 1.25 1.68

The presence of active groups can be used for fixationn of functional agent that provides fiber materials imparting special properties. The assumption of increasing the degree of agent’s fixation on the surface of PET fiber materials after a weak surface hydrolysis was verified in the test experiment. The polyester fabric, treated by sodium hydroxide, put in solution of specifically synthesized phthalocyanine pigment having deodorant action. The test compound has been chosen because of the extremely high coloring, which made it possible to use for the characterization of its adhesion to the fibrous material of the colorimetric analysis methods. The treated sample has a much more intense color than the untreated. It is known that quaternary ammonium compounds (QAC) may be used as phase transfer catalyst-carrier [10-12].Their use can significantly reduce the concentration of alkali in the solution. The catalytic properties of QAC are shown not only in the strongly alkaline, but also in a slightly alkaline medium [11, 12]. The possibility of surface activation of PET material was evaluated by using a dilute solution of sodium hydroxide with the addition of compounds based on QAC using a commercial product of "Ivhimprom", Ivanovo. For the treatment of PET materials by 0.025 M NaOH with the addition of “alkamon OS-2”, “alkamon OS-3”, “alkamon NP” and “triamon” was used. The “triamon” is a cationic compound based on


111

tris(2-hydroxyethyl)methyl ammonium methyl sulfate, the “alkamon” is compositions of cationic (methyl ammonium methyl sulfate derivatives) and non-ionic agents. The QAC concentration was up to 1 g/l. The process was carried out at boiling temperature for 15 minutes. The results at Figure 4 showing that “alkamon” compounds are more effective than “triamon” catalyzing the hydrolysis. Apparently, the nonionic component of alkamons induces dispersing of low molecular weight products (cyclic oligomers of PET [11, 12]) in an aqueous bath. This avoids the blocking of surface active groups of the polymer. The maximum content of surface hydroxyl groups is achieved at a concentration of compounds based on QAC of 0.5 g/l. The breaking load (Table 4) and weight loss (Table 5a) of PET material after the treatment are also measured. It could be noticed from the data presented in Table 4 that the breaking loads are inversely proportional to the concentration of QAS. These data are in consistent with the published data on the mechanism of action QAS [10, 11]. Namely: hydrolysis of fibrous PET material is greatly enhanced in the areas of QAC sorption due to the formation of defects. This phenomenon reduces the strength of the fibrous material reducing its strength. In this case, there is also loss of mass of fibrous material.

Figure 4. The concentration of surface OH groups on the PET fibrous material after treatment by boiling in 0.025 M NaOH with the addition of QAC for 15 min.

Table 4. The breaking load (cN) of PET yarn after treatment by boiling in 0.025 M NaOH with the addition of QAC for 15 min Concentration QAC, g/l Alkamon OS-2 Alkamon OS-3 Alkamon NP Triamon 0.1 1190±101 0.3 1205±100 1150±93 1151±76 1180±83 0.5 1160±82 0.6 1142±93 1141±79 1107±92 0.9 1100±77 1130±93 1070±71 1146±72 The breaking load of PET yarn after treatment by boiling in 0.025 M NaOH is 1210±77 cN

112


Table 5a. The weight loss (%) of PET yarn after treatment by boiling in 0.025 M NaOH with the addition of QAC for 15 min Concentration QAC, g/l 0.3 0.6 0.9

Alkamon OS-2 0.35 1.04 1.21

Alkamon OS-3 0.37 0.54 0.43

Alkamon NP 0.2 0.69 0.88

Triamon 0 0 0

It should be noted that, although the presence of compounds on the basis of QAC forms an additional amount of localized surface hydroxyl groups, while also increasing loss of strength fibrous material. Therefore, it is more favorable for surface activation of PET fiber material while maintaining the original level of strength is handling its solution of sodium hydroxide concentration of 0.125 - 0.25 mol/l at the boiling point for 10 - 15 min. Here is the thread breaking load at baseline. In the presence of compounds based on the number of QAC formed active groups increases, but also a noticeable loss of strength fibrous materials is attained.

3. MODIFICATION OF POLYETHYLENE TEREPHTHALATE FIBERS WITH AQUEOUS SOLUTIONS OF AMMONIA AND AMIDES Strong hydrolysis effect on PET was obtained by the addition of a concentrated aqueous solution of ammonia [13], while, the diluted solutions hydrolyzed the surface localized PET oligomers [14,15]. Consequently, it could be assumed that the presence of an optimum amount of ammonia may provide a weak surface hydrolysis of PET fiber without reducing its strength.

Figure 5. The surface concentration of OH group on the PET fibrous material after treatment by boiling in aqueous solutions of ammonia during: 1 - 10 min. 2 - 15 min. 3 - 20 min.

The PET fiber material was treated in aqueous solutions of ammonia in the concentration range of 0.01-1.15 mol/l at the boiling temperature for 10-20 minutes to determine the


113

optimum process conditions at which the maximum number of hydroxyl groups with no loss on strength. The surface concentrations of OH groups on the treated PET fiber are shown in Figure 5. Comparative analysis of Figure 3 and Figure 5 shows that in the presence of ammonia on the surface of the PET produced more OH groups than in the presence of sodium hydroxide of the same molar concentration, i.e. aqueous ammonia is a more effective catalyst for the hydrolysis of PET material. Figure 4 also shows that an increase in ammonia concentration from 0.1 to 0.9 mol/l the number of formed OH groups almost unchanged. The increase in processing time from 10 to 20 minutes also has little effect on the OH group surface concentration. The presented data in Table 5b showing that changes in PET filament breaking load due to the treatment process using ammonia solutions are within experimental error, i.e. directed hydrolysis does not adversely affect the strength characteristics of the treated yarn. Comparative analysis of the data in Table 1 and Table 5b shows that in the case of processing PET yarns with sodium hydroxide breaking load of the thread remains at the original level by using a solution of the optimal concentration (0.125 - 0.25 mol/l). In the case of treatment by aqueous solutions of ammonia, the strength did not change throughout the investigated concentration range (0.01-1.15 mol/l). Apparently, the presence of ammonia, in contrast to sodium hydroxide, eliminates forming the deeper cracks and micro cavities by surface "etching". Table 5b. The breaking load of PET yarn after treatment by boiling aqueous solutions of ammonia Treatment time, min 0.3 mol/l 0.6 mol/l 10 1123±83 1148±114 15 1184±130 1234±137 20 1161±76 1159±106 Breaking load of untreated yarn is 1180±115 cN

0.9 mol/l 1171±124 1217±79 1156±137

1.2 mol/l 1222±112 1210±139 1192±88

Since ammonia is high fugitive and solutions have sharp unpleasant odor, the alternative processes were suggested [16]. Therefore, in the present study the effectiveness of a carbamide and acetamide as a modifying agents was evaluated. The upper limit of the interval studied concentrations of carbamide (0.33 mol/l) was chosen taking into account the fact that, since the thermal hydrolysis of one molecule of carbamide produced two molecules of ammonia, it corresponds to the concentration of ammonia in a solution of 0.66 mol/l. The next point was the fact that the concentration of ammonia in a solution of ~ 0.1-0.2 mol/l was sufficiently to obtain a high surface concentration of OH group (Figure 5.). Figure 6 shows that OH group surface concentrations are considerably increased after the PET fiber treatment with aqueous solution of carbamide. The high surface concentration of hydroxyl groups is formed after treatment at solution by carbamide concentration of 0.05 - 0.1 mol/l for 15-20 min. Under such conditions the surface concentration of OH ~ 1.4 times more than after treatment with ammonia solutions. Consequently, it could be reported that the mechanism of

114


the intensification of carbamide hydrolysis of PET is not limited to the catalytic effect of ammonia released by carbamide when heated. The results of PET fibers treatment using acetamide and carbamide in different concentrations were presented in Figure 7 a,b. As follows from the data presented in Figure 7 the surface concentration of OH groups on the PET fiber is significantly higher when using carbamide. It can be explained if the thermal decomposition of carbamide lead to not only hydrolysis of the surface of the PET material by ammonia formed. It could be stated that the thermal decomposition of carbamide not leads only to hydrolysis of the PET material surface by formed ammonia but also the aminolysis of PET surface is possibly obtained [13].

Figure 6. The concentration of surface OH groups on the PET fibrous material after treatment by boiling in aqueous solutions of carbamide during: 1 - 10 min. 2 - 15 min. 3 - 20 min.

(a)


115

(b) Figure 7. The concentration of surface OH groups on the PET fibrous material after treatment by boiling in aqueous solutions of carbamide and acetamide during 15 min.: a) vs the initial amides concentration; b) vs the concentration of ammonia formed due to the thermal decomposition of amides.

Figure 8. The results of energy dispersive analysis of PET fabric treated by a carbamide solution of concentration 0.05 mol/l for 15 min.

The JED-2300 Energy Dispersive X-ray Analyzer was used to study the surface of PET fiber material treated with a carbamide solution are shown and the graph was represented in Figure 8. The surface concentration of the elements C, N, and O are measured by this method and presented in Table 6.

116


Table 6. The surface concentration of the elements C, N, and O of PET fabric treated by a carbamide solution of concentration 0.05 mol/l for 15 min element

E,(keV)

At%

Error%

C

0.277

46.17

0.02

N

0.392

29.13

0.34

O

0.525

24.70

0.25

The elemental analysis shows that great surface concentration of nitrogen for treated PET fabric, which confirms our assumption. The most likely mechanism for the formation of surface groups, in our opinion is the exchange reaction of carbamide with PET, and further hydrolysis of obtained products. The measured breaking load of PET fabrics treated with solutions of different concentrations of carbamide are shown in Table 7. The presented breaking load values of PET fibers suggest that the processing with carbamide solutions did not alter its mechanical strength. The comparison of PET fiber surface modifications using sodium hydroxide, ammonia and carbamide shown that the addition of carbamide at concentration of 0.05 - 0.1 mol/l is most effective. Table 7. The breaking load (cN) of PET yarn after treatment by boiling in aqueous solutions of carbamide Treatment 0.05 mol/l 0.017 mol/l time, min 10 1191±120 1152±100 15 1232±104 1177±77 20 1213±114 1203±104 Breaking load of untreated yarn is 1180 ± 115 cN

0.25 mol/l

0.33 mol/l

1246±90 1210±98 1184±137

1208±77 1205±119 1170±101

The presence of reactive groups on the surface of fibrous material and the morphology of surface are great important for technological application. In particular, the adhesion of compounds to a smooth fiber is significantly lower than for rough. In this connection it is interesting to assess the impact of carbamide solution on the morphology of PET fibers. It is known that the treatment of PET materials with solutions of sodium hydroxide lead to "ennobling" provides imparting fiber roughness due to the appearance of etch pits [1,5,17]. The treatment was carried out under hard conditions and accompanied by loss of weight and strength of the material. The images obtained by scanning electron microscopy (Figure 9) shows that PET fibers treated with a solution of carbamide in the optimal mode (concentration of 0.05 mol/l, duration 15 min) are smooth. At the same time atomic force microscopy (Figure 10) shows that on the treated PET film (b) surface structures with lateral length ~ 300 nm were appeared. These structures are apparently associated with a local change in the chemical composition of the surface.


a

117

b

Figure 9. The scanning electron microscopy images of PET fibers treated with 0.05 M carbamide solution during 15 min at boiling point. Zoom: a – x1100; b – x5000.

Figure 10. Phase contrast images of PET film (size 5x5 μm) (AFM): a - initial, b - treated with a solution of carbamide concentration of 0.05 mol/l for 15 min

4. EFFECT OF SURFACE MODIFICATION OF PET FABRICS FOR THEIR ANTIMICROBIAL FINISH The hygienic protection of human is one of important task for medicine, transport, military condition etc. One way to perform this target is developing of textile materials with biological activity [18]. With the high demand in the sale of a limited range of such products made from natural fabrics, though, for many reasons, it would be better using of PET fiber materials. The latter of relatively low cost, have the durability, dimensional stability, high aesthetic characteristics and are widely used for clothing and linen sports and medical applications [13]. PET materials have a good resistance to microorganisms [13], but some types of bacteria can grow on the impurities that may appear on the surface. One way to improve the antimicrobial properties of materials is fixing on the surface of a special compound which imparts the required properties on the surface. However, for PET based textile materials this approach is very difficult, because there are a very small number of active groups on the

118


surface of PET fibers. Increasing the surface concentration of active groups is need for suitable fixation of antimicrobial agents on the PET surface. Above we have shown that in case of PET fibrous material the formation of chemically active groups in the PET surface layer can be achieved by controlled hydrolysis. In our opinion, we can expand the range of compositions for antimicrobial finishing of PET materials due to oxygen-containing reactive groups formed on the fiber surface. The good fixation leads to improve the quality and durability antimicrobial finish also [19]. Preparation of antimicrobial finishing of PET fabric was selected from among of the nonmigratory, since biocide product must only be applied to the substrate (textile material), and do not migrate to human skin, causing allergic reactions to its action [20]. The most popular product are non-migratory Sanitized T99-19 and T25-25 Sanitized - Silver, produced by Swiss company Sanitized AG (Sanitized AG), which is a world leader in the development and production of biocide agents for textile, leather, paper and plastics. Sanitized T25-25 - Silver (chloride of silver as active substance) is traditionally used for antimicrobial finishing PET fabrics [20]. It is fixed on fabrics due to intermolecular interaction. As alternative the low cost finishing agent Sanitized T99-19 was considered for the permanent finishing of cellulose fabrics. This composition is based on the quaternary ammonium compound of silicon and leads to antimicrobial and antifungal protection. It is not harmful to humans and the environment and has high levels of protection against microorganisms and bacteria. For antimicrobial finishing the PET fabric surface were pre-treated by 0.05 M carbamide solution during 15 min at boiling point. Then the sample was washed by distilled water and dried in air. After that the fabric was modified in an aqueous solution of antimicrobial agent (concentration around 0.4% by weight of fabric) during 5 minutes at boiling temperature. Then the modified sample was washed by distilled water and dried in air again. The biocide effect of Sanitized T99-19 modified PET fabric was studied in the Ivanovo State Medical Academy. The Gram-positive and Gram-negative bacterial cultures: Staphylococcus aureus (S. aureus - Gram(+)) and Escherichia coli strain M-17 (E. coli Gram(-)) was tested. The antimicrobial finishing stability after washing and dry friction [8] was evaluated also. At the first stage the inhibition of microorganisms growth (lysis zone) around the samples after 24 hours incubation at 37ºC was measured. In this test we observed the absence of inhibition zone around the samples. It should be noted that it is typical for non-migratory agent, which not diffused into the culture medium. At the second stage the simplified version of the counting of microbiological test ASTM E 2149 [21] was applied. This test is based on counting the number of colonies produced into saline solution after 24 hour contacting with grinded modified PET fabric. Initial number of colonies was inserted as suspension. The number of colonies formed was evaluated photometrically due to changes in turbidity (light scattering) of the solution. The results of Sanitized T99-19 application for antimicrobial finishing of PET fabrics are given in Table 8. According to the results presented in Table 8 the initial fabric after antimicrobial finishing by Sanitized T99-19 is not demonstrate antimicrobial effect. The absence of microbial activity of the initial fabric after antimicrobial finishing shows that Sanitized T99-19 is not fixed on inactivated surface. At the same time the fabric pre-treated by 0.05 M carbamide solution during 15 min after antimicrobial finishing demonstrate excellent microbial protection. Thus, it was found that pretreatment process allows using Sanitized T99-19 for antimicrobial finishing of PET fabric. Biocide effect is great effective


119

against Gram-positive bacteria (Table 8). The antimicrobial effect achieved is highly resistant to washing and relatively to dry friction. Table 8. ASTM E 2149 test of Sanitized T99-19 application for antimicrobial finishing of PET fabrics State of PET fabric Initial fabric after antimicrobial finishing fabric pre-treated by 0.05 M carbamide solution during 15 min after antimicrobial finishing fabric pre-treated by 0.05 M carbamide solution during 15 min after antimicrobial finishing and dry friction test fabric pre-treated by 0.05 M carbamide solution during 15 min after antimicrobial finishing and washing test

colonies growth, % E. coli S. aureus 28.5 21.5 0

0

6.6

1.3

1

0

CONCLUSION In this paper the special properties of textile materials was obtained due to surface fixation of a special compounds which imparts the required properties. In the case of PET the fixation of compounds are difficult because there are practically absence the chemically active groups on the surface. To perform this problem the surface localized hydrolysis using various chemical reagents as process initiator was investigated. It was confirmed that hydrolysis of the surface layer of fibrous PET materials leads to carboxyl and hydroxyl groups formation. By comparative analysis of the concentration of surface OH groups after treatments was shown that the carbamide solutions at concentration of 0.05 - 0.1 mol/l was most effective for modification. These conditions were suitable for formation of oxygen- and nitrogen containing groups’ on the surface without noticeable changes in PET filament breaking load. The pretreatment was applied for antimicrobial finishing of PET fabrics using low cost non-migratory agent Sanitized T99-19. Excellent biocide effect against Grampositive and Gram-negative bacterial cultures of finished fabric was stated. The textile material properties were resistant to washing and relatively to dry friction.

REFERENCES [1] [2]

[3]

Bruce M. Latta. Improved Tactile and Sorption Properties of Polyester Fabrics through Caustic Treatment // Text. Res. J., 1984. V. 54. № 11. P. 766–775. Gorodnichaya T.YU., Kovtun L.G., Krichevskii G.Ye., Trofimov N.A. Nekotorye zakonomernosti shchelochnogo gidroliza polietilentereftalatnykh tekstilʹnykh materialov // Izvestiya VUZov. Tekhnol. tekst. prom-sti. 1989. T 4. S. 74 – 77. Rudakova T.Ye., Moisyeev YU.V., Chalykh A.Ye., Zaikov G.Ye. Kinetika i mekhanizm gidroliza polietilentereftalata v vodnykh rastvorakh gidroksida kaliya // Vysokomolek. soedineniya. 1972. Ser. A. T. 14. № 2. S. 449 – 453.

120 [4] [5]

[6]

[7]

[8] [9] [10]

[11]

[12] [13] [14] [15]

[16]

[17] [18]

[19]

[20]

[21]

N. P. Prorokova, A. V. Chorev, S. M. Kuzmin et al. Alkalisierung: gewichtsabbau auf Polyester // Chemiefas. – Textilind. 1989. V. 39. N 5. S. 475. Kabaev M.M., Pashkyavichus V.V., Darzhininkai‘tite O.V., Libonas YU.YU. Vliyanie shchelochnoi‘ obrabotki na stroenie pripoverkhnostnykh sloev elementarnykh nityeĭ iz polietilentereftalata // Khimich. volokna. 1988. № 5. S. 52 – 53. Sadov, F.I., Sokolova, N.M., Vil’dt Ye.O. i dr. Laboratornyi praktikum po kursu khimicheskaya tekhnologiya voloknistykh materialov / pod red. prof. F.I. Sadova. M.: GizLegProm, 1963g. S. 127-129. Baskakova, T.I. Volkova, L.Ye., Glazkovskii’, YU.V. i dr. Analiticheskii kontrol proizvodstva sinteticheskikh volokon: Spravochnoe posobie / pod red. A.S. Chegoli i N.M. Kvasha. M.: Khimiya, 1982. S. 142–143. Krasiteli dlya tekstilnoi’ promyshlennosti // pod red. A.L. Byal'skogo, V.V. Karpova. M.: Khimiya, 1971. 312 s. Dekhant I., Dants R., Kimmer V., Shmol'ke R. Infrakrasnaya spektroskopiya polimerov / pod red. E.F. Olyei’nika. M.: Khimiya, 1976. 472 s. Gorodnichaya T.YU., Tsarevskaya I.YU., Kovtun L.G., Krichevskii G.Ye. Issledovanie intensifitsiruyushchego vozdyei’stviya organicheskikh spirtov i chetvertichnykh ammonievykh soedinenii na protsess shchelochnogo gidroliza polietilentereftalatnogo volokna // Izvestiya VUZov. Tekhnol. tekst. prom-sti, 1990. № 5. S. 61–65. Prorokova N.P., Vavilova S.YU., Prorokov V.N. Vozdyei’stvie na polietilentereftalatnoe volokno preparatov na osnove chetvertichnykh ammonievykh soedinenii // Khimich. volokna. 2007. № 6. S. 17 – 20. Prorokova N.P., Vavilova S.YU., Prorokov V.N. Vozdyei’stvie solyei ammoniya na polietilentereftalatnye materialy // Khimich. volokna. 2007. № 1. S. 17 – 22. Petukhov V.V. Poliefipnye volokna. M.: Khimiya, 1976. 272 s. Vavilova S.YU., Prorokova N.P., Kalinnikov YU.A. Vliyanie vodnykh rastvorov ammiaka na polietilentereftalatnoe volokno // Tekst. khimiya, 1995. № 2 (7). S. 70–77. Prorokova N.P., Vavilova S.YU. Modifitsiruyushchyee dyei’stvie slabykh vodnykh rastvorov ammiaka na polietilentereftalatnye materialy // Khimicheskie volokna, 2006. №6. S. 15–18. Prorokova N.P. Vozdyei’stvie na oligomery polietilentereftalata kak sposob intensifikatsii periodicheskogo krasheniya poliefirnogo volokna // Izvestiya VUZov. Khimiya i khim. tekhnologiya, 2007. T. 50. Vyp. 3. S. 57–62. Zaikov G.Ye., Razumovskii S.D. Destruktsiya kak metod modifikatsii polimernykh izdelii // Vysokomolek. soedineniya, 1981. Ser. A. T. 23. № 3. S. 513–531. Razuvaev A.V. Biotsidnaya otdelka tekstil'nykh materialov. Chast 1 // Rynok legkoi’ promyshlennosti, 2009. №60. Elektronnyi resurs: http://www.rustm.net/catalog/ article/1453.html Bossard M. Gigienicheskaya zashchita tekstilnykh materialov kak argument dlya prodazhi izdelii. Primer vysokogo marketinga // Ros. Khim. zhurnal: ZH. Ros. Khim. Ob’va im. D. I. Mendelyeeva, 2002. t. XLVI. №2. – S. 62–65. Razuvaev A.V. Zaklyuchitelnaya otdelka tekstil'nykh materialov biotsidnymi preparatami // Izvestiya VUZov. Khimiya i khim. tekhnologiya, 2010. T. 53. Vyp. 8. S. 3–7. ASTM E2149 - 10 Standard Test Method for Determining the Antimicrobial Activity of Immobilized Antimicrobial Agents under Dynamic Contact Conditions. USA, 2001.



MODELING THE MOVEMENT OF DUST PARTICLES IN THE SWIRLING FLOW R. R. Usmanova1 and G. E. Zaikov2 1

Ufa State Technical University of Aviation, Ufa, Bashkortostan, Russia 2 N.M. Emanuel Institute of Biochemical Physics, Russian academy of sciences, Moscow, Russia

ABSTRACT An important part of mathematical modeling of the cleaning gas scrubber is the study of the trajectories of the individual particles of dust and study the dependence of the trajectories of the most significant factors. Calculated trajectories of particles . This will find the conditions for trapping particles and quantify the impact of relevant factors on the process of gas purification. Calculated complexes that characterize the geometry and the scrubber operating parameters.

Keywords: the trajectory of particles, equation of motion of a particle, the coefficient of resistance, geometric complex, custodial complex

1. INTRODUCTION To studying vortex devices in many papers, extensive experimental data. However, many important problems of analysis and design of vortex devices have not yet found a systematic review. Existing research in this area show a strong sensitivity of the output characteristics of the regime and design of the device. This indicates a qualitatively different flow hydrodynamics at different values of routine-design parameters. Thus, it is important to consider the efficiency of fluid flow and vortex devices, receipt and compilation dependencies between regime-design parameters of the machine. Creating effective designs is the actual problem.



E-mail:[email protected]

122

R. R. Usmanova and G. E. Zaikov

2. STATEMENT OF THE PROBLEM, THE ASSUMPTIONS One of the most common devices for dust cleaning equipment considered centrifugal machines. Due to their widespread use simplicity of design, reliability, and low capital cost. Consider the mechanism of dust-gas cleaning scrubber is something dynamic [1]. Capture dust scrubber is based on the use of centrifugal force. Dust particle flows at high velocity tangentially enters the cylindrical part of the body and makes a downward spiral.The centrifugal force caused by the rotational motion flow, dust particles are moved to the sides of the device (Figure 1). When moving in a rotating curved gas flow dust under the influence of centrifugal force and resistance. Analysis of the swirling dust and gas flow in the scrubber will be carried out under the following assumptions: 1. Gas considered ideal incompressible fluid and, therefore, its potential movement. 2. Gas flow is axisymmetric and stationary. 3. Circumferential component of the velocity of the gas changes in law

w  const  r This law is observed in the experiments [2, 3], will provide a simple solution that is convenient for the quantitative analysis of the particles. 4. Particle does not change its shape over time and diameter, it does not happen any crushing or coagulation. Deviation of the particle shape on the field is taken into account the coefficient K. 5. Wrapping a strong flow of gas is viscous in nature. Turbulent fluctuations of gas is not taken into account, which is consistent with the conclusion of [4]: turbulent diffusion of the particles has no significant impact on the process of dust removal. 6. Not considered force Zhukovsky, Archimedes, severity, since these forces by orders of magnitude smaller than the drag force and centrifugal [5, 6, 3]. 7. Concentration dust is small, so we can not consider the interaction of the particles 8. Neglect the uneven distribution of the axial projection of the radial velocity of the gas, which is in accordance with the data of [7]. Axial component of the velocity of the particles changes little on the tube radius. The rotation of the purified stream scrubber creates a field of inertial forces, which leads to the separation of a mixture of gases and particles. Therefore, to calculate the trajectories of the particles it is necessary to know their equations of motion and aerodynamics of the gas flow. In accordance with the assumption of a low concentration of dust particles on the influence of the gas flow can be neglected. Consequently, we can consider the motion of a single particle in the velocity field of the gas flow. Therefore, the task to determine the trajectories of particles in the scrubber is decomposed into two parts:  

Determination of the velocity field of the gas flow; Integration of the equations of motion of a particle for a calculation of the velocity field of the gas.

Modeling the Movement of Dust Particles in the Swirling Flow

123

Figure 1. The trajectory of the particles in a dynamic scrubber.

The assumption of axial symmetry of the problem (with the exception of the mouth) allows for consideration of the motion of the particles using a cylindrical coordinate system. The greatest difficulty is to capture fine dust, for which the strength of the resistance with sufficient accuracy is given by Stokes. By increasing the ratio of dust cleaning machine grows [8], so the calculation parameters scrubber with low dust content (by assumption 6) guarantees a minimum efficiency.

3. DERIVE THE EQUATION OF MOTION OF A PARTICLE To calculate the trajectories of the particles need to know their equations of motion. Such a problem for some particular case is solved by the author [9]. We introduce a system of coordinates OXYZ. Its axis is directed along the OZ axis of symmetry scrubber (Figure 2). Law of motion of dust particles in the fixed coordinate system OXYZ can be written as follows:

m

d v 'p dt

 Fst

where m - mass of the particle;

d v 'p - velocity of the particle; Fst - aerodynamic force.

(1)

124


For the calculations necessary to present the vector equation (1) motion in scalar form. Position of the particle will be given by its cylindrical coordinates (r; φ; z). Velocity of a particle is defined by three components: Uр-tangential, Vр - radial and Wр - axial velocity.

Figure 2. The velocity vector of the particle.

We take a coordinate system O'X'Y'Z ', let O'X' passes axis through the particle itself, and the axis O'Z' lies on the axis OZ. Adopted reference system moves forward along the axis OZ Wр speed and rotates around an angular velocity

(2)

1 The equation of motion of a particle of mass m   p d 3p coordinate system O'X'Y'Z' 6 becomes:

m

d v 'p dt

        

 Fst  ma0'  m rp'    m rp'    m   rp'    2m v p  

or

d vp dt



     

1 Fst  a0'  rp'      rp'    2 v p   m

'

where a0 -translational acceleration vector of the reference frame; d v p -velocity of the particle;

(3)


125

rp' -the radius vector of the particle;

r   - acceleration due to unevenness of rotation; ' p

  r    - the centrifugal acceleration; ' p





2 v p   - Coriolis acceleration. The first term on the right-hand side of equation (3) is the force acting c gas flow on the particle, and is given by Stokes



Fst  3 g d p v g  v p



(4)

 g - dynamic viscosity of the gas. The second term (3) is defined as

dW p

ez 

dt

dW p dt

ez'

Convert the remaining terms:

r     r  ddt   r  dtd  Ur p



  r

1 p

' p



 



U p2





p

p

  1 dU p U p  '  dU p U pV p  e y ez '    rp    2 V x e y      dt   r dt  r r p p     p 

e  e  e    r e  e   r U p2

z

rx

' p

x

z

U p2

z

v

p

ex

p

 U pV p  U e y ' 2 v p    2v x ' e x '    2 v x ' x e x '  e z '    2  rp rp   













where eч , ev , ez – vectors of the reference frame and used the fact that rp  e x  rч  v x  V p Substituting these expressions in the equation of motion (3),

m or

d v 'p dt

        

 Fst  ma0'  m rp'    m rp'    m   rp'    2m v p  

126


dvp dt



     

1 Fst  a0'  rp'      rp'    2 v p   m

We write this equation in the projections on the axes of the coordinate system O'X'Y'Z'

 dVx ' 1 U p2  F   stx ' m rp  dt  dU p U pV p 1   0  Fsty  dt rp m   dW p 0  1 Fstz  m dt  or

 dV p 1 U p2 F    stx m rp  dt  dU U pV p  p 1  Fsty   m rp  dt  dW  p  1 Fstz m  dt

(5)

We have the equation of motion of a particle in a rotating gas flow projected on the axis of the cylindrical coordinate system. Substituting (2) and (4) in (5) we obtain the system of equations of motion of the particle:

 dV p U p2 18      V V  p  p d p2 g rp  dt  dU  p 18 U g  U p U pV p   2 ч dч rp  dt  dW  p  182 Wg  W p   pd p  dt

(6)

4. OUTPUT RELATIONSHIP BETWEEN THE GEOMETRICAL AND OPERATIONAL PARAMETERS. Formal analysis of relationships that define the motion of gas and solids in the scrubber. The analysis shows that the strict observance of similarity of movements in the devices of different sizes requires the preservation of four dimensionless complexes, such as


Re d 

wD



; Fr 

127

 2 w2 v ; Ar  ; Re   Dg D1 

Not all of these systems are affecting the motion of dust. Experimentally found that the influence of the Froude number Fr negligible [10] and can be neglected. It is also clear that the effect of the Reynolds number for large values it is also insignificant. However, maintaining unchanged the remaining two complexes, still introduces significant difficulties in modeling devices. On the other hand, there is no need for strict observance of the similarity in the trajectory of the particle in the apparatus. What is important is the end result - providing the necessary efficiency unit. To estimate the parameters that characterize the removal of particles of a given diameter, consider the approximate solution of the problem of the motion of a solid particle in a scrubber. A complete solution for a special case considered in the literature [11], this solution can be used to obtain the dependence of the simplified model of the flow. For the three coordinates - radial, tangential and vertical equations of motion of a particle at a constant resistance can be written in the following form: 2

dw w    wr  v r  dt r dwz   wz  v z  dt dw w  2 wr   w  v  dt r where α-factor resistance to motion of a particle, divided by its mass.



 K 2

K-factor, which takes into account the effect of particle shape (take K = 2). The axial component of the velocity of gas and particles are the same, as follows from the equations of motion by neglecting gravity. Indeed, if the

dwz   wz  v z  , then taking dt

dwz  wz Wz  wzo e  dt dt If initially wz  v z Wz  0 ; Wt  0 и Wt  const projection speed:

wz  v z  wz , we get

w  const  r Valid law Wφ(r) may differ markedly from the accepted, but this is not essential. In this case, it only makes us enter into the calculation of average

128


 w2   r 

    av

Under these simplifications, the first of the equations of motion is solved in quadrature. Indeed, for now

w2 dwr  awr  r dt then with the obvious boundary condition t = 0, vr = 0, we have

1  w   r

2

wr 



av 1  e

 dt

   



The time during which the flow passes from the blade to the swirler exit from the apparatus as well

t1 

l wz av

On the other hand, knowing the law of the radial velocity, we can find the time during which the particle travels a distance of r1 (the maximal distance from the wall) to the vessel wall (r2).

r2  r1   wr dt  0

r2  r1 

w dr

 1  e dt  dt

0

v  l 1 dt  t1  e  1  rav   

Substituting in this equation is the limiting value of t1 = tt. we get:

rav r2  r1  v2 av or

 l  l 1   2  wz e    1  wz   

Modeling the Movement of Dust Particles in the Swirling Flow 

r12  r22 wz K 2 wz     1 e      K 2 2V2 av l l  

2

l  K wt

  1  

129

(7)

The presented approach is based on the known dependence and model of the flow, it's different in a number of studies approach is only in the details. However, further to allocate two sets, one of which characterizes the geometry of the device, and the other - operating data. The use of these systems simplifies the calculation and, most importantly, takes into account the influence of some key factors to the desired gas velocity and height of the apparatus WZav. Dependence structure (7) shows the feasibility of introducing two sets, one of which

A

l 2  2Wt K

characterizes the effect of the flow regime and the particle diameter, and the other is a geometric characteristic of the device.

Ar 

r2 l

1  r ctg 2 1

(8)

In (8) through r1 marked relative internal radius apparatus:

r1  r1 / r2 and β1 - the average angle of the flow at the exit of the guide apparatus

tg 1 

V Vz

Then (7) takes the simple form

Ar  f  A where

f ( A) 





1 1 2 A  l 1 A 2 A2

(9)

130


Expressed graphically in Figure 3, this dependence allows to determine the minimum value of the mode parameter Amin for the scrubber with geometrical parameters of Ar. And must take A >Amin One of the important consequences of the resulting function is the relationship between the diameter of the dust particles and the axial velocity of the gas. With this machine, with data A g, Amin = const and therefore w  z

2

 const

0,5 0,45 0,4 0,35 0,3 0,25 0,2 0,15 0,1 0,05 0

10; 0,45 Аг

f(А)

1000; 0,2 4000; 0,1 10000; 0,05 0

5000

10000

15000; 0 15000 20000

Operational parameters Ар

Figure 3. The relationship between the geometrical and operational complexes.

This means that reducing the particle size axis (expenditure), the rate should be increased according to the dependence

 o    w zo      wz

2

Unfortunately, a significant increase Wz permitted as this may lead to the capture of dust from the walls and ash. You can also change the twist angle β and the height of the flow system, without changing the axial velocity. If the reduction of the particle diameter δ or increasing the size of the unit has increased the value of A, it must be modified accordingly Ag (using a graph Figure 3), and the new value of Ar to find an angle β: ctg ctg

0

 Ag A  go



   


131

CONCLUSION 1. Creating a mathematical model of the motion of a particle of dust in the swirling flow allowed us to estimate the influence of various factors on the collection efficiency of dust in the offices of the centrifugal type, and create a methodology to assess the effectiveness of scrubber. 2. Identified settlement complexes, one of which characterizes the geometry of the scrubber and the other operational parameters. The use of these systems and simplifies the calculation takes into account the influence of several key factors. 3. The developed method can be used in the calculation and design of gas cleaning devices, as constituent relations define the relationship between the technological characteristics of the dust collectors and their geometrical and operational parameters.

REFERENCES [1] [2] [3] [4]

Pat. RF 2339435 ,2008. G.M. Barahtenko, I.E. Idelchik Industrial and sanitation gas. 1974. 6., 4-7. V.Straus Industrial cleaning gases. Chemistry, Moscow, 1981, 616 p. M.I. Shilyaev aerodynamics and heat and mass transfer of gas-dispersion flow: studies. allowance. Publishing house of Tomsk. State. architect. - Builds. University, Tomsk, 2003. 272 p. [5] M.G. Lagutkin, D.A. Tohti sheep, 2004.38, 1, 9-13. [6] M.E. Deutsch, G.A. Filippov, Gas Dynamics of two-phase media. Energy, Moscow, 1968, 424 p. [7] A.V. Starchenko, A.M. Bells, E.S. Burlutskiy Thermophysics and Aeromechanics, 1999.6, 1, 59-70. [8] Gupta, D. Lilly, N. Sayred swirling flow Trans. from English. Ed. SY Krashennikova. Mir, Moscow, 1987, 588 p. [9] D.A. Bezic. Diss.kand.tehn. sciences, gos.inzh.-tehnol.akademiya Bryansk, Bryansk, 2000.150p. [10] V.E. Mizonov, V. Blaschek, R. Colin, A. Greeks Tohti, 1994.28, 3, 277-280. [11] A.T. Litvinov, Journal of Applied Chemistry, 1971.44, 6, 1221-1231.